Intermittent fasting: eating window and timer

In trials — about the same. Systematic reviews of randomised trials compared intermittent fasting with continuous calorie restriction and found comparable weight loss; where a difference appeared it was small and often clinically insignificant. The point of fasting is not a special metabolic effect but the fact that some people find it easier not to eat at all than to eat a little and police themselves constantly. If that is you, the schedule helps. If you simply eat the same amount in eight hours instead of sixteen, the weight will not move.

That is the most repeated number in the topic and there is no human data under it. Fasting-induced autophagy has been studied in detail in rodents; in humans it is measured rarely and indirectly, through markers in leukocytes and muscle tissue, and the timing is unsettled enough that a dedicated trial about it is running right now. That is why autophagy on this tool’s scale is labelled “mostly from mice” and sits past the 24-hour mark rather than at hour sixteen. All that can be said with confidence: the longer the gap, the more reason to expect the effect — but “it switched on at 16:00” cannot be promised.

On the available data, earlier is better. In trials of an early window, where eating finished in the middle of the day, insulin sensitivity and blood pressure improved even without weight loss, while late meals are handled worse by glucose control and make falling asleep harder. This is the one part of the topic where the placement of the window changes something on its own, which is why the tool checks the closing time against your bedtime and warns when fewer than three hours separate them. Practical upshot: if the choice is between skipping breakfast and skipping dinner, the data leans towards the latter.

With 14:10 or 16:8. The 16:8 protocol is the most common and usually amounts to skipping breakfast or an early dinner — many people already live close to it without calling it fasting. Longer variants such as 20:4 and OMAD narrow the window sharply, fitting a normal amount of food and protein into it becomes difficult, and slips get more frequent. Starting with long gaps for speed is the usual way to quit after two weeks.

Calories. Water, black coffee, unsweetened tea, electrolytes and creatine do not break it; coffee with milk, juice, broth, alcohol, protein powder, BCAAs and a fish-oil capsule do, even in small amounts. Calorie-free sweeteners technically do not break anything, but in some people they sharpen appetite and make holding out harder. The most common unnoticed break is the splash of milk in coffee: scientifically it changes almost nothing, but the habit of adding it tends to grow. The “What you can drink” tab holds 28 entries with a search box.

You can, if you feel fine. Light and moderate work is tolerated normally on an empty stomach, whereas heavy resistance training at the end of a long gap often goes worse. It is more practical to place the session close to the eating window so that protein arrives within the following hours. If a fasted session makes you dizzy, that is not “adaptation” — it is a reason to eat.

Usually because the window limits time, not calories. Eight hours is plenty to eat a full day’s intake and more, especially when the first meal lands on heavy hunger. Check the amount rather than the schedule: count calories for a couple of days without changing your habits. The second common cause is caloric drinks inside the window — a latte and a juice do not register as food but count in full. The third is short windows like 20:4, after which the volume eaten in one sitting grows more than it feels.

No, it works on a different principle. In 5:2 the restriction runs by days: five days of normal eating and two days of roughly 500–600 kcal. Hour-based protocols such as 16:8 limit time rather than amount, which is why they require no calorie counting. This tool is built for hour-based protocols; 5:2 needs a calendar and a calorie target for the restricted day, not a countdown. Mixing them into one timer would mean showing a number that measures nothing.