You Started Cutting and the Scale Went Up. Here Is Why
A 500 calorie deficit removes about 65 grams of fat a day. Everything else the scale weighs moves by more than that — here is what, and how much.
Written while building GETSIXPACK — a training log that scores the session against your range and calls the next load.
You are nine days into a cut. You have weighed everything, you have not touched the biscuits, you have trained four times, and this morning the scale read eight hundred grams more than it did on Monday.
Nothing has gone wrong. The scale is doing its job perfectly — it is weighing everything you are made of, and only one of those things is the thing you are trying to lose.
What a day of fat loss actually weighs
Start with the arithmetic, because it reframes the entire problem.
A kilogram of body fat is worth roughly 7,700 calories. Run a 500 calorie daily deficit — which is the ceiling, not the target — and you are removing about 65 grams of fat a day. The weight of an egg.
That is the whole story. Everything below is a list of the things that move by more than an egg’s worth in a day, so you can stop reading them as failure.
Glycogen, and the water that travels with it
Muscle and liver store carbohydrate as glycogen, and glycogen is stored wet.
Fernández-Elías and colleagues took muscle biopsies from nine trained cyclists before and after 150 minutes of riding in the heat, and again through recovery. Exercise dropped muscle glycogen by 44% and muscle water by 13%. During recovery, in the condition where fluid was deliberately under-replaced, every gram of glycogen restored came back with three grams of water. Their conclusion is about as flat as sports science gets: each gram of glycogen is stored in human muscle with at least three grams of water.
Now watch what that does to a diet.
Yang and Van Itallie ran six people through three ten-day periods each and measured what the lost weight was made of. On an 800 calorie ketogenic diet they lost 466.6 grams a day. On an isocaloric mixed diet — the same 800 calories, with carbohydrate in it — they lost 277.9 grams a day.
The faster diet did not remove more fat. It removed more water.
This runs in both directions and that is the useful part. It is why cutting carbohydrate looks like it is working spectacularly in week one. It is why the scale jumps up two days after a normal weekend. And it is why the week you eat at maintenance during a diet break shows a gain that comes straight back off the week after.
Being fair to the evidence: a 2022 review of how muscle glycogen is measured notes that at the whole-body level the relationship between stored glycogen and total body water is still not settled. The biopsy work is about muscle. What follows from it for the number on your bathroom scale is an inference — a well-founded one, and an inference.
Training puts water in too
If you have just started a programme, or come back to one, or changed it, part of what you are carrying is swelling.
Damas and colleagues followed ten untrained young men through ten weeks of resistance training with ultrasound. At week three, vastus lateralis cross-sectional area was up about 2.7% — and so was the echo intensity of the image, alongside elevated myoglobin and interleukin-6 in the blood. Their conclusion is a warning about measurement: early increases in muscle size are not purely hypertrophy, because oedema from muscle damage is sitting in there too, and it may account for a large proportion of the change.
That was measured on a thigh, not on a scale, so do not put a kilogram figure on it. What it tells you is that a muscle that has just been trained hard holds fluid it did not hold last week — and every part of you that does that is on the scale in the morning.
Salt is the usual suspect, and the case is weaker than you think
Everyone knows the salty-meal story. The evidence for it, when somebody actually controlled the conditions, is oddly thin.
Heer and colleagues put 32 healthy men in a metabolic ward and ran their sodium chloride intake at 50, 200, 400 and 550 milliequivalents a day. Plasma volume rose dose-dependently — by 315 mL at the highest intake. Total body water did not increase. Body mass did not increase either. What appears to happen is a shift of fluid from the interstitial space into the bloodstream rather than a net gain of water.
It is not the last word — a later study found the extracellular fluid response to high sodium depends on BMI, so the picture is not the same in everybody. But if you had a big restaurant meal and the scale was up a kilogram the next morning, the sodium is a smaller part of that than the obvious things: the food itself has mass, so do the drinks, and neither of them has left yet.
That is worth saying plainly because it is the least mystical item on this page. A day’s food and fluid weighs more than a day’s fat loss by an order of magnitude, and it stays inside you for a while.
The week has a shape, and you are probably sampling one point of it
Orsama and colleagues analysed 4,657 daily weights from eighty adults, each tracked for between fifteen and 330 days, and found a group-level weekly rhythm that is worth knowing about before you set your weigh-in day.
Weight is highest early in the week — Sunday and Monday. It falls from Tuesday. It starts rising again on Saturday. The rise-and-fall cycle repeats, and the people whose weekday compensation was strongest were the ones who lost or maintained over the long run; the weakest compensators were the slow gainers.
The authors put the practical conclusion in the paper itself: weekend-to-weekday variation should be treated as normal rather than as a sign of gain. If you weigh once a week on a fixed day, you are not tracking your bodyweight — you are tracking one phase of a repeating wave, and any two consecutive readings differ partly because of where the wave was.
And if you menstruate, about half a kilogram of it is scheduled
Kanellakis and colleagues measured 42 women twice a week across a full cycle. Bodyweight was significantly higher during menstruation than in the first week of the cycle, by 0.450 kg — and the increase was accounted for by a statistically significant rise of 0.474 kg in extracellular water. Nothing else about body composition changed significantly.
Half a kilogram, entirely fluid, arriving on a schedule. Against a signal of 65 grams a day, that is a week of apparent progress appearing or disappearing for reasons that have nothing to do with the diet.
So read the trend, and give it three weeks
None of this makes the scale a bad instrument. It makes a single reading a bad measurement, which is a different problem with a boring solution.
- Weigh every morning, under the same conditions. After the bathroom, before food and drink, same clothing or none.
- Average the seven readings and compare averages, never days. A weekly average against last week's weekly average is a signal. Tuesday against Monday is not.
- Ignore the first two weeks entirely for estimating your rate. That is where the glycogen and its water leave, and it will flatter you.
- Give any change three weeks before acting on it. Two flat weekly averages in a row is a stall. One is a Tuesday.
- Keep a second instrument. Waist at the navel, weekly, same morning. It moves for fewer reasons than bodyweight does.
- Watch your top sets. Strength falling across several lifts for two weeks is a real signal about the deficit, and it is not made of water.
There is one more reason the trend is the only thing worth reading, and it is not about noise at all. Hall and colleagues modelled adult weight change for The Lancet and the central property of the model is that the body’s response to a change in intake is slow: expenditure keeps falling as you get lighter, so the deficit you set in week one is quietly closing the whole time. A weight trend is the only instrument you own that can see that happen. A single reading, however carefully taken, cannot.
If the scale has been flat for nine days and you have started wondering whether the whole thing is working, the honest answer is that nine days is not enough data to know — and that the number of weeks this is actually going to take is worth calculating now rather than finding out by attrition.
Where this comes from
Muscle biopsy work confirming at least three grams of water stored per gram of glycogen: Fernández-Elías, Ortega, Nelson and Mora-Rodríguez, European Journal of Applied Physiology, 2015 (doi:10.1007/s00421-015-3175-z). The caveat that the whole-body relationship remains unsettled: Shiose, Takahashi and Yamada, Nutrients, 2022 (doi:10.3390/nu15010155).
The composition of weight lost on ketogenic versus mixed 800 calorie diets — 61.2% water against 37.1% — is Yang and Van Itallie, Journal of Clinical Investigation, 1976 (doi:10.1172/JCI108519).
Early training-induced increases in muscle size being partly oedema rather than hypertrophy: Damas, Phillips, Lixandrão, Vechin and colleagues, European Journal of Applied Physiology, 2016 (doi:10.1007/s00421-015-3243-4).
High sodium intake raising plasma volume without raising total body water or body mass: Heer, Baisch, Kropp and Gerzer, American Journal of Physiology — Renal Physiology, 2000 (doi:10.1152/ajprenal.2000.278.4.F585). The BMI-dependence of the extracellular fluid response: Visser, Krikken, Muntinga, Dierckx and colleagues, Obesity, 2009 (doi:10.1038/oby.2009.61).
The weekly weight rhythm across 4,657 daily measurements: Orsama, Mattila, Ermes and van Gils, Obesity Facts, 2014 (doi:10.1159/000356147).
The 0.450 kg rise at menstruation, accounted for by extracellular water: Kanellakis, Skoufas, Simitsopoulou, Migdanis and colleagues, American Journal of Human Biology, 2023 (doi:10.1002/ajhb.23951).
Why a static calculation drifts and the trend is the only instrument that sees it: Hall, Sacks, Chandramohan, Chow and colleagues, The Lancet, 2011 (doi:10.1016/S0140-6736(11)60812-X).
Questions people ask about this
- Why did I gain weight in the first week of my diet?
- Almost certainly not fat. A 500 calorie daily deficit is worth about 65 g of fat a day, and the scale is also weighing your glycogen and the water bound to it, your gut contents, the fluid your training put into the muscle, and where you are in the week. Any one of those moves by more in a day than the fat does.
- How much can bodyweight change in a day without any fat changing?
- More than the fat does, comfortably. Two of the sources have been measured directly: extracellular water accounts for a rise of about 0.47 kg at menstruation, and bodyweight follows a repeating weekly pattern that peaks on Sunday and Monday. Add glycogen, gut contents and training fluid on top of that.
- Should I weigh myself every day?
- Yes, and you should ignore every individual reading. Weigh under the same conditions each morning, average the seven, and compare the averages. One number is a sample of the noise; a weekly average is a measurement of the trend, and it is the only instrument that can see your maintenance calories moving during the cut.
- Does salty food make you gain weight?
- Less reliably than the folklore says. When researchers took 32 healthy men in a metabolic ward from 50 to 550 milliequivalents of sodium a day, plasma volume rose dose-dependently — but total body water did not increase and neither did body mass. What a salty meal out actually puts on the scale the next morning is mostly the meal, the drinks, and what has not left yet.