One meal spends part of a week, not all of it#
A single large meal cannot undo a week of dieting, and the arithmetic is easy enough to do at the table. Suppose you are running a 500-calorie daily deficit; that is 3,500 calories banked over seven days. Suppose the meal in question runs to 2,000 calories where your plan had 700. That overshoot is 1,300 calories — roughly 37% of the week's deficit, which turns a seven-day loss into about a four-and-a-half-day one. (That worked example is mine, not a figure from any study.) To actually erase the week you would need to eat about 3,500 calories above plan in one sitting, which is a physical undertaking rather than a slip.
So the reassurance is real. What makes this topic worth more than a paragraph is that the reassurance rarely lands, because the scale disagrees the next morning — sometimes by a kilogram or two — and because the measurable damage from a cheat meal has nothing to do with the calories in it. The best-evidenced cost is downstream of the label, and it has been isolated in an experiment where the calories were held constant and the forbidden-ness was varied.
Why the morning after weighs so much more than you ate#
A large mixed meal is heavy before it is anything else: food, fluid and the salt that pulls water with it are all sitting inside you. But the part that surprises people is glycogen. A big carbohydrate load refills muscle and liver glycogen, and glycogen is warehoused wet — which is why the number can jump far beyond anything the energy content could possibly deposit.
The capacity of that buffer has been measured, in an extreme and very small experiment. Three men had their glycogen stripped by three days of low-carbohydrate eating plus exercise, then were overfed carbohydrate for seven days at 3,642 kcal/day rising to 4,930 kcal/day. Glycogen storage capacity came out at roughly 15 g per kilogram of bodyweight, and the stores could absorb a gain of about 500 g of glycogen before net lipid synthesis began contributing to body fat1. Three subjects is three subjects, and this was a deliberately punishing protocol; treat the number as an order of magnitude rather than a personal allowance. The shape of the finding is what matters: there is a real, sizeable buffer between a large carbohydrate meal and your fat stores, and filling it registers on the scale as several times its own weight in water. That reading is ordinary fluid movement, and it clears over the following days.
The surplus does become fat — just not the way the internet says#
Here is where the comforting version of this article usually overreaches. "Carbs don't turn into fat" is repeated as though a large meal were free. The measurement says something more specific and less flattering.
Eight lean and five obese women were overfed by 50% for 96 hours with either sucrose or glucose, with de novo lipogenesis tracked by deuterium incorporation into plasma triacylglycerols and fat balance measured by whole-body calorimetry. Lipogenesis rose two- to threefold with overfeeding — and in absolute terms ran from about 2 g/day at energy balance to a maximum of 10 g/day when overfed. Over the same 96 hours, mean fat balance was approximately +275 g2. The authors' conclusion is that manufacturing fat from carbohydrate "does not contribute greatly to total fat balance."
Read both numbers together. The body barely bothers converting carbohydrate into fat; it simply burns the carbohydrate you just ate and stops burning the fat you ate alongside it, which goes to storage instead. The fat still arrives. It arrives in proportion to the surplus, by a boring route, at roughly the rate your excess energy implies — which is exactly why the arithmetic in the first paragraph is the right instrument and the biochemistry is a distraction.
| The claim | What the measurement says |
|---|---|
| "One big meal makes you fat overnight" | No — the overnight jump is food, fluid and glycogen's bound water |
| "Carbohydrate turns straight into body fat" | Rarely — de novo lipogenesis maxed at ~10 g/day even under 96 h of 50% overfeeding |
| "So a cheat meal is free" | No — fat balance still ran +275 g over those 96 hours, via dietary fat spared from oxidation |
| "One meal can undo a week" | Only if it exceeds plan by your entire weekly deficit — around 3,500 kcal on a 500/day cut |
The expensive part is the word#
Now the finding this whole topic ought to be built on, because it separates the energy from the label cleanly. In a three-study series, restrained eaters were given preloads and then covertly measured on how much they ate afterwards. The classic milkshake preload was compared against non-forbidden preloads of equivalent calories. The result: food type, not perceived calories, was the element of the preload that caused disinhibition — and in the third study, only restrained eaters anticipating a forbidden food were disinhibited, whether that food was high or low in calories3.
The calories did not trigger the collapse. The category did — and low-calorie forbidden food triggered it just as well.
That is a 1989 laboratory paradigm in restrained eaters, not a field study of dieters at a restaurant, and it should be read at that strength. But it makes a testable prediction about the modern practice, and the modern practice bears it out. A 2025 scoping review of the cheat-meal literature — eight studies, searched to October 2024 — found that when relaxed eating was framed as goal-directed, it was associated with positive effects on eating behaviour including reduced hunger and greater satisfaction, whereas framing the same food as "contradictory to one's goals" or as a reward correlated with disordered eating patterns. Evidence for the physiological claims people make about cheat meals — lean-mass retention, blunted metabolic adaptation, better training — came out "mixed"4.
And the cultural version is larger than the arithmetic at the top of this page assumes. A thematic analysis of 600 images drawn from over 1.6 million #cheatmeal posts found 54.5% depicting very large quantities and 71.3% depicting calorie-dense foods, at volumes the authors rated as qualifying for an objective binge episode5. A 1,300-calorie overshoot is a Friday dinner. What the word has come to advertise is several times that, photographed.
What to do the day after, and what not to#
The damage-control instinct is the actual risk, because it is the same instinct the preload studies were measuring. Three things follow from the evidence above:
- Do not repay it. Skipping the next day's meals or adding a punishing session converts one overshoot into a restriction-and-rebound loop, and the laboratory version of that loop is triggered by the sense of violation rather than by the energy. Return to your normal intake, not to a smaller one.
- Do not read the scale for a few days. You have added food, fluid and glycogen water; a fortnight's trend is the only reading that can tell you what the meal actually cost, and the week is the accounting unit anyway.
- Budget it in advance if you can. A meal you subtracted from somewhere is a plan; the same meal unsubtracted is a story about willpower. If you want the version with a protocol and a small evidence base behind it, that is a refeed rather than a cheat.
- Watch frequency, not magnitude. One 1,300-calorie overshoot costs about a third of a week's deficit. Three of them costs the week, and then the arithmetic in the opening paragraph stops being reassuring and starts being the diagnosis.
The framing worth keeping: a large meal is a withdrawal from a balance you have been building, and balances survive withdrawals. What they do not survive is a bookkeeping system that treats one withdrawal as bankruptcy — which is a matter of how you hold the tracking rather than what you ate, and is the only part of this that has ever ended a diet.
FAQ#
How much would you have to eat in one meal to undo a week's deficit?#
About as much as the whole week's deficit, in one sitting. On a 500-calorie daily cut that is 3,500 calories above your plan for that meal — not 3,500 calories total, but 3,500 more than the meal was supposed to be. A large restaurant dinner running 1,300 calories over plan spends roughly a third of the week, which is the arithmetic behind why single meals feel more consequential than they are.
Should I fast or add cardio the day after a big meal?#
Neither is required, and both carry a cost the meal did not. Compensatory restriction is the trigger side of the restrained-eating cycle — in laboratory work, disinhibited eating followed from a food being classed as forbidden rather than from its calories3. Going back to your ordinary intake settles the matter arithmetically, since one overshoot is a fraction of a week and the week is what your body integrates.
Is calling it a "cheat" meal actually a problem?#
The framing has better evidence behind it than the physiology does. A scoping review of eight studies found relaxed eating framed as goal-directed was linked to reduced hunger and greater satisfaction, while framing it as rule-breaking or as a reward correlated with disordered eating4. Content analysis of #cheatmeal posts found most depicted binge-sized portions of calorie-dense food5. Calling it a planned high-calorie meal costs nothing and changes what it licenses.
Sources#
- Acheson KJ, Schutz Y, Bessard T, Anantharaman K, Flatt JP, Jéquier E. Glycogen storage capacity and de novo lipogenesis during massive carbohydrate overfeeding in man. Am J Clin Nutr. 1988;48(2):240-247.
- McDevitt RM, Bott SJ, Harding M, Coward WA, Bluck LJ, Prentice AM. De novo lipogenesis during controlled overfeeding with sucrose or glucose in lean and obese women. Am J Clin Nutr. 2001;74(6):737-746.
- Knight LJ, Boland FJ. Restrained eating: an experimental disentanglement of the disinhibiting variables of perceived calories and food type. J Abnorm Psychol. 1989;98(4):412-420.
- Tsang JH, Poon ETC, Trexler ET, Wong SH, Zheng C, Sun F. The Role of Cheat Meals in Dieting: A Scoping Review of Physiological and Psychological Responses. Nutr Rev. 2025;83(11):2240-2252.
- Pila E, Mond JM, Griffiths S, Mitchison D, Murray SB. A thematic content analysis of #cheatmeal images on social media: Characterizing an emerging dietary trend. Int J Eat Disord. 2017;50(6):698-706.
