The plate is empty. You set the fork down, lean back, and the verdict arrives from somewhere under your ribs: full. Not stuffed. Just finished. And then, maybe ninety seconds later, a second voice speaks up, quieter and more persuasive than the first. Something sweet. Not a meal. Just a little something. You are not hungry. You know you are not hungry. You open the cupboard anyway.

If that sequence feels like a small nightly failure, here is the uncomfortable part: it is not a failure of willpower, and it is not really about hunger at all. Your stomach was consulted, and it answered honestly. Your mouth was never asked. It has opinions of its own, and it waits until the stomach has finished talking before it offers them.

Fullness is not one thing

We speak about being full as if the body kept a single gauge, a tank that reads empty or full. It doesn't. It tracks at least two different kinds of enough.

The first is the one you expect: stretch receptors in the stomach wall, and gut hormones such as cholecystokinin, GLP-1 and peptide YY that rise as food arrives in the intestine. This signal is general. It is about volume and energy, and it does not much care what the food tasted like.

The second is stranger, and it has a name: sensory-specific satiety. In the early 1980s, Barbara Rolls and her colleagues at Oxford ran a simple series of experiments. People rated how pleasant a range of foods were, ate one of them until they were satisfied, then rated everything again. The food they had just eaten dropped sharply in pleasantness. The foods they had not eaten barely moved. The decline was specific to the sensory properties of what had been consumed: its taste, its texture, its smell.

The effect is strongest across categories. After a savoury meal, your own tongue has quietly rated savoury things down. Sweet was never on the plate, so sweet has never been rated at all. That channel is still wide open, pleasantness intact, waiting.

That is the dessert stomach. It is not a second compartment. It is a second channel. And it opens precisely when the first one closes.

Why the craving arrives on schedule

There is a reasonable evolutionary story here. An animal that stopped eating the moment any one food became boring, but stayed interested in foods it had not yet tried, would end up with a varied diet and a wider range of nutrients. Sensory-specific satiety is the mechanism behind the well-documented variety effect: offer people more distinct flavours in one sitting and they eat more. Buffets are built on it. Dessert is simply the oldest buffet, reduced to a single extra course.

Layer a learned cue on top of that biology and the timing becomes almost mechanical. For most of your life, the end of a savoury meal has been followed by something sweet. Birthdays, restaurants, your grandmother's kitchen, the mints at the till. The moment of the fork going down has been paired with sweetness thousands of times. The brain is a prediction machine, and the craving you feel is a prediction, not a need. It says: this is usually when the sweet thing comes.

Recent research hints the two signals may be closer neighbours than we assumed. In 2025, a group at the Max Planck Institute for Metabolism Research in Cologne reported that in mice, the same POMC neurons that signal fullness also release an opioid-like signal when sugar is available, switching on an appetite for sugar in animals that had just eaten to satiety. Mice are not people, and the human picture is still being worked out. But the shape of it matches what you feel at the table: the system that says enough and the system that says but sweet are not opponents. They may share an address.

Where this meets your fasting window

For anyone doing intermittent fasting, after dinner is the hinge of the entire day. The last bite sets the clock. Dinner at half past seven followed by something sweet at twenty to ten does not add a small indulgence to your evening. It moves the start of your fast by more than two hours, and it does so with the least attentive eating of the day: standing up, from the container, half-watching something.

That is the real problem with the after-dinner sweet, and it is not the sugar. It is that the meal never ended. It trailed off. A dinner that trails off has no last bite, only a series of candidates, and the fast cannot begin until the series does.

Meals need endings

Notice how many food cultures solved this long before anyone measured a gut hormone. Mint tea after a Moroccan meal. A sliver of paan. A single espresso. A piece of fruit, cut and shared. These are not second courses. They are punctuation. They tell the tongue, in a flavour it recognises, that the sentence is over.

There is a psychological reason endings matter beyond habit. Work by Daniel Kahneman, Barbara Fredrickson and Donald Redelmeier on what became known as the peak-end rule showed that our memory of an experience leans heavily on its most intense moment and on how it ended, far more than on its length. A meal that ends with a stale biscuit eaten over the sink is remembered as that biscuit. It feels unfinished, slightly disappointing, and an unfinished meal is one the brain is happy to reopen an hour later. A meal that ends on purpose, with something small and deliberate, is filed as complete. Complete meals do not nag.

So the goal is not to defeat the sweet craving. The goal is to give dinner an ending the craving can accept, and then to actually stop.

Your next moves

  • Put the sweet on the plate, not after it. A few slices of fruit, a square of dark chocolate, a spoon of sweetened yoghurt, served alongside dinner tonight. It arrives while the savoury channel is still open, so it joins the meal instead of restarting it. The portion you put on a plate is almost always smaller than the one you take from a packet.
  • Choose a non-food ending and perform it at the exact moment you would normally reach for dessert. Mint tea, brushing your teeth, running the dishwasher and switching the kitchen light off. The cue is the fork going down, and the cue needs a new response delivered at that instant, not ten minutes later when you are already standing at the cupboard.
  • Decide about dessert before you sit down, never after. If you want it, plan it, plate it, and put the rest away before the first bite of dinner. A decision made full, cued and tired is a worse decision than the same one made in advance.
  • Leave the table for ten minutes before you judge the craving. Change rooms, do one small task, then check again. Sensory-specific satiety keeps the sweet channel open, but the urge itself is largely cue-driven and it fades when the cue is out of sight. Treat this as a measurement, not a test of character.
  • For one week, write down the time of your genuine last bite. Not calories, just the clock. Compare it to when dinner actually finished. The gap is the tax the trailing-off is charging your fast, and seeing the number is usually enough to make you want to close it.

The last bite is the one that counts

Upvas is built on a simple idea: your eating window should fit around your dinner, not the other way round. Dinner is the anchor, and the fast begins when dinner truly ends. When you log that last bite, you see exactly where your window started, and ending the meal on purpose turns out to be the easiest way to lengthen your fast without changing a single thing you eat. If your evenings have been trailing off instead of finishing, give them a proper ending and let the clock start honestly at upvas.lumenlabs.works.