Intermittent Fasting Works About as Well as Eating Less, Because in the Trials That Is What It Turned Out to Be

Alabama, 2022. The most favourable result came from the University of Alabama at Birmingham, where 90 adults with obesity were given weight-loss counselling and randomised to an early eating window, 7am to 3pm, or to a window of twelve hours or more. Over fourteen weeks the early eating group lost 2.3 kilograms more, 6.3 against 4.0, a significant difference. But the additional loss was not fat; body fat and the ratio of fat to weight lost were no different between the groups. Early eaters also reported better mood and less fatigue, which may be the more useful finding.

What the pattern says

Put together, the trials show that time-restricted eating works when, and to the extent that, it causes people to eat less. In TREAT it did not cause them to eat less and it did not work. In Guangzhou everybody was already eating less and the window added nothing. In Alabama the early window helped people stick to the counselling and they lost a little more weight, though not more fat. There is no demonstrated metabolic bonus in humans beyond the calorie reduction. The switch that impressed everyone in mice has not yet shown up as extra benefit in a person.

That is not a failure. Any pattern that reliably helps someone eat less without misery is valuable, and for some people a clock is easier to obey than a calorie count. Skipping a late evening of grazing is a genuine and common way that the window does its work. The point is that it is one such tool, not a different kind of tool.

The scare in the other direction

In March 2024 a conference abstract presented to an American Heart Association meeting reported that people who ate within an eight-hour window had a 91 per cent higher rate of cardiovascular death. It was widely reported and widely misunderstood. It was an observational analysis based on two days of self-reported eating recall per participant, it had not been peer reviewed or published in full at the time, and people who eat in a short window for reasons of illness, shift work or poverty look very different from people who choose to. It is not evidence that time-restricted eating is dangerous. It is evidence that dietary recall surveys are a poor way to study it.

Who should not

• Anyone with type 1 diabetes, or type 2 diabetes treated with insulin or sulfonylureas, without medical supervision; the hypoglycaemia risk is real. Our guide to fasting risk categories in diabetes covers the medical framework.
• Anyone pregnant or breastfeeding.
• Anyone with a current or past eating disorder, for whom rules about when food is permitted are a known trigger.
• Children and adolescents.
• People on GLP-1 medication who are already struggling to eat enough protein, a problem set out in our piece on muscle loss on these drugs.

For a healthy adult who finds it easy, an eating window is a reasonable way to eat less. For one who finds it hard, it offers nothing that eating less some other way would not, and the trials give no reason to persist with it.