Eating late does change your metabolism. Whether it makes you fatter is less clear.

Eating the same food later in the day is not metabolically identical to eating it earlier. In a randomised crossover trial that held calories, nutrients, physical activity, sleep and light exposure constant, late eating increased hunger, altered appetite-regulating hormones, reduced waking energy expenditure and shifted gene expression in fat tissue in a direction consistent with less fat breakdown and more fat storage [s1]. What has not been demonstrated is that this translates into meaningful weight gain in ordinary life: the observational literature finds a 17% higher odds of obesity among late eaters, an association whose direction of causation it cannot establish [s2].

The trial, and why its design matters

Most studies of meal timing cannot separate when people eat from what and how much they eat, and from how they sleep. This one tried to. Participants with overweight and obesity completed both arms of a randomised, controlled crossover in which nutrient intake, physical activity, sleep and light exposure were rigorously controlled, with the only difference being the timing of the meals [s1].

Late eating increased hunger (P < 0.0001) [s1]. It altered appetite-regulating hormones, raising both the waking and the 24-hour ghrelin-to-leptin ratio (P < 0.0001 and P = 0.006 respectively) [s1]. It decreased waking energy expenditure (P = 0.002) and 24-hour core body temperature (P = 0.019) [s1]. And adipose tissue gene expression analysis showed changes in lipid metabolism pathways — p38 MAPK signalling, TGF-β signalling, modulation of receptor tyrosine kinases and autophagy — in a direction consistent with decreased lipolysis and increased adipogenesis [s1].

That is a coherent mechanistic package: hungrier, burning slightly less, and with fat tissue tilted toward storage. The authors describe it as converging mechanisms by which late eating may result in positive energy balance and increased obesity risk [s1] — a conditional, and deliberately so. The trial measured mechanisms, not weight change.

The population data, and what they cannot settle

A 2026 meta-analysis pooled three case-control and five cohort studies of late eating and obesity risk, searching to September 2025 [s2]. Across all included studies, late eating was significantly associated with higher odds of obesity (pooled odds and hazard ratio 1.17, 95% CI 1.03 to 1.34) [s2].

The subgroup breakdown is where the confidence drains out. The association was significant in cohort studies (1.20, 1.04 to 1.38), which were mostly conducted in Asian populations, and absent in case-control studies (0.86, 0.49 to 1.50) [s2]. It appeared in adults and in high-quality studies, but not in adolescents or in low-quality ones [s2]. Eight studies is a small evidence base, and the authors call for further prospective cohorts to confirm the findings in adolescents and in Western populations [s2].

Observational meal-timing data also carry a particular confounding problem. People who eat late are disproportionately shift workers, people with irregular sleep, people who skip earlier meals, and people whose evenings involve alcohol or takeaway food — all independently associated with weight. No adjustment fully removes that.

What is established and what is not

Established: eating late alters hunger signalling, energy expenditure and adipose gene expression under controlled isocaloric conditions [s1]. Associated: later eating tracks with higher obesity odds in adults in cohort studies, at a pooled 1.17 [s2]. Not established: that shifting meals earlier causes weight loss, which would require a randomised trial with weight as an outcome and a duration measured in months rather than days. Neither of these papers is that trial.

The size of the effect is also worth keeping in proportion. A 1.17 odds ratio is a modest association, and the trial's mechanistic findings — a measurable drop in waking energy expenditure, a shifted hormone ratio — were detected under laboratory conditions designed to maximise sensitivity, not in free-living people [s1].

The thermoregulatory finding, and why it is interesting

One of the trial's secondary results is easy to skip past. Late eating reduced 24-hour core body temperature (P = 0.019) alongside the drop in waking energy expenditure (P = 0.002) [s1]. Core temperature is one of the most stable circadian outputs the body produces, and moving it by shifting meal times — with sleep and light held constant — is evidence that food timing is acting as a circadian input in its own right, not merely as a proxy for the sleep schedule that usually travels with it.

That is the strongest argument the trial makes, and it is a mechanistic one. It says the body treats a late meal differently from an early one. It does not say by how much over a year, in a person who is not in a metabolic ward.

Where the trial's population limits it

The crossover trial enrolled adults with overweight and obesity [s1]. Whether the same hormonal and adipose responses appear in people of normal weight is not answered by it. Nor does the trial address the version of this question most people are actually asking — whether a late dinner in the context of an otherwise stable diet matters — because the trial shifted the entire eating window rather than adding a meal at the end of one. The meta-analysis, meanwhile, is built on eight observational studies with heterogeneous definitions of what counts as "late" [s2].

This article is informational and is not medical advice. Meal timing interacts with diabetes medication schedules, reflux and shift-work sleep disorders, and those are clinical questions.

Sources

• Late isocaloric eating increases hunger, decreases energy expenditure, and modifies metabolic pathways in adults with overweight and obesity — Cell Metabolism, 2022
• Late Eating and Obesity: A Meta-Analysis of Observational Studies — Nutrition Reviews, 2026