"If it's up, I ate too much yesterday"

The number is up, and you know why, because you can remember last night. The curry, the second glass, the bread you didn't need. The rise feels like the evidence and the meal feels like the crime. It's a neat story and it's the wrong one.

The meal is still in you. Whatever you ate after six last night hasn't all left, and what has left has been partly replaced by the water the salt in it is holding and the water the carbohydrate in it is now stored with. Kreitzman and colleagues make the point in the paper from chapter 3 that when someone who's been on a very low calorie diet eats carbohydrate again, the regain looks exaggerated, because the fuel store refills and brings its water with it. That's the extreme version. The mild version is Saturday night's rice showing up on Sunday's scale. The rise is real. It's also mostly leaving.

There's a study that shows the shape of this at the scale of a week. Anna-Leena Orsama and colleagues took 4,657 daily weights from 80 adults over periods of two weeks to nearly a year and found a pattern. Weight was highest on Sunday and Monday and fell through the week. The rise began on Saturday and the fall began on Tuesday. The people who lost weight or held steady over the whole period were the ones with the strongest weekday fall, and the authors' conclusion was that the weekend-to-weekday swing should be treated as normal rather than as a sign of gaining. I'd add something they don't say, which is that a Monday reading taken as a judgement on Sunday's dinner will be wrong every single week.

Look at your sheet from chapter 1. Find the mornings you called bad, and look at the evening before each one. My guess is that most of them follow a bigger, saltier or later meal, and that the number came back down within two days whether or not you did anything about it. If that's what your sheet shows, you've just watched noise, and you've watched yourself settle up for it.

What a rise on one morning tells you about fat is, on its own, nothing. What a rise that stays for three weeks tells you is something, and chapter 16 is about how to tell the two apart. The mistake isn't caring about the second kind. It's reading the first kind as though it were the second, every morning, and feeling the swing each time.