Meal timing didn't beat usual care on liver fat

Time-restricted eating has become a default suggestion for weight and metabolic health, and a persistent claim is that when you eat matters as much as how much — with early windows supposedly better aligned to circadian biology. Liver fat is a good test of that idea, because it responds quickly to energy balance and is measurable precisely by MRI. If eating-window timing has independent effects, an early window should outperform a late one. This trial set out to compare early, late and self-selected windows against standard dietary care.

In a multicentre randomised trial, adults with overweight or obesity, half of them women, were assigned to usual care alone (49), early time-restricted eating (49), late time-restricted eating (52) or self-selected time-restricted eating (47) for 12 weeks. Usual care was a Mediterranean diet-based education programme, and the time-restricted groups added an 8-hour eating window on top of it. Hepatic fat fraction was measured by MRI, with liver health assessed via elastography-based parameters, liver enzymes and circulating biomarkers. Faecal microbiota was profiled by 16S rRNA gene sequencing.

Hepatic fat fraction fell significantly within all three time-restricted groups, but none differed from usual care: mean differences were -0.4% for early, -1.5% for late and -0.7% for self-selected windows, and the timing groups did not differ from each other. Liver health markers and faecal microbiota showed no between-group differences either. Participants who already had metabolic dysfunction-associated steatotic liver disease at baseline, and those who lost at least 5% of body weight, had substantially larger reductions in liver fat (mean differences -2.7% and -2.6%). More people reached that ≥5% threshold on time-restricted eating than on usual care, at 41-44% versus 16%.

This complicates the popular framing that early eating windows carry a special metabolic advantage, at least for liver fat over three months. It reframes time-restricted eating as a tool that helps some people achieve clinically meaningful weight loss rather than as a timing intervention with independent effects on the liver. The null on microbiota is notable given how often eating windows are linked to gut changes. The authors caution the trial was not powered for these secondary outcomes, and 12 weeks may be too short for slower-moving liver markers.

This is a reason to pick an eating pattern you can actually stick to rather than agonising over whether to eat earlier or later in the day. The liver fat reductions in this trial travelled with weight loss and with how much fat was in the liver to begin with, not with the clock. It also suggests that a defined window can help more people reach a meaningful weight-loss threshold, which may be its real value. None of this is a treatment plan, and the trial only followed people for 12 weeks.