What the schedules are
- 16:8 — sixteen hours without food, an eight-hour eating window, usually noon to eight in the evening. Since sleep covers eight of the sixteen, the practical change is skipping breakfast or dinner. This is the most studied of the time-restricted patterns.
- 18:6 and 20:4 — narrower windows. Total intake tends to fall further, partly because there is less time to eat, and each meal has to carry more. Anyone training for strength has to fit their protein into the window in fewer sittings, which gets awkward past about six hours.
- OMAD — one meal a day. Simple to describe and hard to do well: fitting a day's protein, fibre and micronutrients into a single sitting is genuinely difficult, and it is the pattern most associated with rebound overeating.
- 5:2 — not a time window at all. Five days of normal eating, two non-consecutive days at 500 to 600 calories. Different mechanism, different failure mode, and the reason the first-meal input is ignored when you select it.
What the trials show
The honest summary is that time-restricted eating produces weight loss when it reduces total intake, and not otherwise. Two studies are worth knowing about. TREAT, published in JAMA Internal Medicine in 2020, randomised participants to 16:8 or three structured meals for twelve weeks and found the difference in weight loss was small and not statistically significant, with a signal suggesting some of the loss in the fasting group was lean mass. A 2022 trial in the New England Journal of Medicine, running for a year in China, compared calorie restriction with a time-restricted window against the same calorie restriction without one and found no additional benefit from the window.
That does not make it useless. For some people, a rule about when to eat is easier to follow than a rule about how much, and adherence is the thing that determines whether any dietary change works. But the mechanism is behavioural. Claims that fasting windows unlock autophagy or reprogram metabolism in humans rest largely on animal work and on cell culture, at fasting durations well beyond sixteen hours, and the human evidence for clinically meaningful effects is thin.
Choosing the window
The window that survives contact with your life is the right one. If dinner with other people is not negotiable, a one in the afternoon to nine in the evening window is the workable choice. If you cannot function without breakfast, eight to four is fine.
Circadian research does suggest earlier windows have a small advantage for glucose and blood pressure markers, since insulin sensitivity is higher in the morning. The effect is modest, and a late window you keep will beat an early window you abandon by a wide margin. Treat earlier as a tiebreaker, not a requirement.
During the fasting window the standard is water, black coffee and unsweetened tea. Milk, sugar, juice and anything with calories end the fast in the strict sense. If your goal is reduced total intake, a splash of milk in coffee is not going to matter. If you believe you are chasing a specific metabolic state, be aware that the human evidence for that state existing at these durations is weak.
Who should not do this
The medication issue is the serious one and it is easy to overlook. Insulin and sulfonylureas lower blood glucose whether or not you have eaten, so skipping meals on an unadjusted dose risks hypoglycaemia, which can be severe. Many other drugs are specified to be taken with food, either for absorption or to avoid gastric irritation, and a fasting window quietly makes that impossible. Neither problem is solved by being careful; both require the prescriber to adjust something.
Beyond that: pregnancy and breastfeeding, any history of an eating disorder, being underweight, and adolescence are all reasons not to. Restrictive eating schedules can reactivate disordered patterns in people with a history of them, and the rule-based structure that makes fasting appealing is precisely what makes it risky in that group.
Signals to stop
Shaking, cold sweats or severe dizziness during a fast suggest hypoglycaemia and mean stopping and eating, then working out why it happened before trying again. Menstrual irregularity or amenorrhoea, hair loss, worsening sleep and a growing preoccupation with food are all reasons to abandon the schedule rather than push through it. Persistent fatigue that does not settle after two or three weeks of adaptation is a sign the pattern does not suit you, and there is no prize for continuing.
Questions people ask
Can I exercise during the fasting window?
Low and moderate intensity aerobic work is generally fine and many people prefer it. Hard strength training is a different question: performance in the session tends to be lower without prior food, and protein availability afterward matters for the adaptation. The practical arrangement is to place heavy sessions at the end of the fast so that the eating window opens shortly after, which gets you the meal within an hour or two of finishing. If you become lightheaded during fasted training, that is information rather than something to work around.
Does a splash of milk in coffee break the fast?
Strictly, yes — anything with calories does. Whether it matters depends entirely on what you are after. For weight management the mechanism is reduced total intake, and twenty calories is noise against a daily total. For the metabolic states people invoke, such as autophagy, the human evidence is weak enough that optimising around a splash of milk is optimising against something that has not been demonstrated. Artificial sweeteners are debated and the data on their effect on insulin response in humans is mixed and mostly small-scale.
Will I lose muscle on a fasting schedule?
You can, and the TREAT trial found a signal suggesting a meaningful share of the weight lost in the fasting arm was lean mass. The risk rises as the window narrows, because total protein intake tends to fall and because fewer feeding occasions means fewer stimuli for muscle protein synthesis. The countermeasures are ordinary ones: keep protein intake up, around 1.6 g per kilogram of body weight for someone training, spread it across the meals you do have, and keep doing resistance training. A four-hour window makes all three harder.
Is 16:8 better than just eating less?
The randomised evidence says no, it is a different way of arriving at the same place. The 2022 NEJM trial is the clearest test: everyone restricted calories, half also restricted the window, and the window added nothing over a year. What varies between people is which rule they can actually keep. Some find a clock easier than a calorie count and do better on it; others eat more in eight hours than they previously ate in fourteen. The right question is not which method is superior in trials but which one you will still be doing in six months.
Why does the 5:2 option ignore the first meal time?
Because 5:2 restricts by day, not by hour. On a restricted day you eat 500 to 600 calories whenever suits you, and on the other five days there is no timing constraint at all. It is a fundamentally different intervention from time-restricted eating that happens to sit under the same heading, and applying an eating window to it would be inventing a rule that is not part of the protocol. The two restricted days should be non-consecutive, which is the only scheduling detail 5:2 actually specifies.