How to Come Off a Calorie Deficit on Purpose
What do I actually do when I want to stop losing weight?
Raise your intake to the maintenance figure for your new weight — not to whatever you were eating before the diet started — and keep weighing yourself. That is the whole transition: one arithmetic step, plus the discipline to treat maintenance as a target with a number on it rather than as the period where you stop having a number.
The reason people regain is almost never that they chose maintenance. It is that they never chose anything. The deficit just stopped being observed, intake drifted back to a level that suited a heavier body, and the scale did what arithmetic said it would.
How do I work out my maintenance number?
You already have the data. If you logged reasonably honestly through the last stretch of the diet, your own results tell you more than any online calculator will.
Take your average intake over the last six to eight weeks and your average rate of loss over the same period. A kilogram of body fat is about 7,700 kcal, so:
Average intake: 1,850 kcal a day
Average loss: 0.35 kg a week
0.35 × 7,700 = about 2,700 kcal a week, or roughly 385 kcal a day of deficit
Estimated maintenance: about 2,200–2,300 kcal a day
Use a range, not a point. Your logging was an estimate and your weight trend was noisy, so the answer is a band. Round it and start at the middle. The longer method, including how to handle a window where your weight barely moved, is in How to Find Your Real Maintenance Calories.
Now compare that to what you ate before you started. Someone who was eating 2,600 kcal at 12 kg heavier and goes straight back to 2,600 is running a surplus of around 350 kcal a day. That is roughly 2,450 kcal a week, or about a third of a kilogram of fat every two weeks — not dramatic in any single week, which is exactly why it goes unnoticed for months. A lighter body burns less at rest and costs less to carry up stairs; maintenance falls as you lose, and the old number is now the wrong number.
Should I add the calories back all at once or gradually?
Either works, and I would not pretend the evidence supports a precise protocol. You will find confident schedules online — add 100 kcal a week, hold for a fortnight, repeat — presented as established physiology. The honest version is that the destination is what matters and people differ in how they get there comfortably.
Going straight to the estimated maintenance figure is simplest and gives you a clean signal: whatever the scale does over the next month is the answer to whether the number was right.
Adding roughly 100–150 kcal a day each week for two or three weeks is gentler if a sudden jump in food volume makes you uneasy, or if you want the scale's step to be smaller and less alarming. It also takes longer to find out whether you guessed well.
What matters either way is that the extra calories come from food you would eat indefinitely. Two hundred extra calories as a proper breakfast addition is a habit. Two hundred extra calories as "I'm at maintenance now" is not a number at all.
Why has my weight gone up 1.5 kg in a week?
Because you refilled the tank. This is the part nobody warns people about, and it ends more maintenance attempts than genuine regain does.
Eating more — particularly more carbohydrate — restores muscle and liver glycogen, and glycogen is stored with several times its own weight in water. More food eaten means more food physically in transit through the gut at any moment. Sodium intake usually rises with portion sizes, and sodium holds water too.
Added together, an increase of roughly 1 to 2 kg in the first one to two weeks is ordinary, and none of it is fat. To put a boundary on it: gaining 1.5 kg of actual fat requires a surplus of about 11,500 kcal. Nobody does that in five days by adding a bowl of porridge. The same mechanics behind a one-morning jump are covered in Why Your Weight Went Up Overnight.
How do I tell that step apart from real regain?
By shape, not size.
The water-and-glycogen step happens once. It rises over roughly one to two weeks and then flattens. Draw a line through your daily weigh-ins and it looks like a stair: up, then level.
Fat regain does not flatten. It carries on at a steady slope for as long as the surplus does. So give the transition a fortnight, ignore the scale's absolute number during it, and then judge the next three to four weeks on the trend alone. If your weekly average is still climbing by something like 0.2–0.3 kg a week a month after the step settled, that is a surplus, and it is arithmetic you can fix by trimming 200–300 kcal a day.
If it is level — even at a number higher than your lowest weigh-in — you are maintaining. Your lowest weigh-in was a partly dehydrated version of you and was never going to be the settling point.
How much should I still log at maintenance?
More than nothing. The failure mode of maintenance is not a collapse, it is drift: portions creeping up by 10% across the board. A tablespoon of olive oil is about 120 kcal and a second one is invisible on the plate; USDA FoodData Central will tell you the same figure whether or not you noticed pouring it.
Three workable levels:
Keep logging fully. Least interesting option, most accurate, and if logging is already a five-minute habit there is no reason to abandon a tool that works.
Log a few days a week. Two weekdays and one weekend day a week catches most drift. Weekends are usually where the arithmetic hides.
Log nothing, weigh daily. Cheapest, and viable — provided you have decided in advance what the scale is allowed to do before you resume logging.
Option three only works with a defined band, which is the single most useful thing to set up now rather than later.
What should my weight band be?
Write down two numbers before you need them, because you will negotiate with yourself in the moment.
Pick your settling weight — the flat average after the first fortnight, not your lowest morning ever. Then set an upper trigger somewhere around 2 kg above it, and a rule for what happens when the weekly average, not a single morning, sits above it for two consecutive weeks: you go back to full logging for a fortnight and find out where the extra energy is.
A fortnight of tracking is a small, finite task. Discovering the same thing 8 kg later is not. The patience required is the same patience described in How Long a Weight Loss Stall Has to Last Before It Means Anything, applied in the other direction: two weeks of noise means nothing, four weeks of slope means something.
A lower trigger is worth having too. If you are still losing at maintenance, your estimate was low and you should raise it.
One thing worth being clear about
Maintenance is not a reward phase and it is not the end of the job — it is the longer job. It is also, for most people, the easier one, because the target is bigger and a single heavy meal barely registers against it.
If you are using Day So Far, the target it learns keeps working at maintenance for the same reason it works in a deficit: it compares what you logged against what the scale actually did over a fortnight, so the figure it settles on is a measurement of your new body rather than a formula's guess about your old one. Whatever you use, the requirement is the same — a number, a band, and enough attention to notice when the trend stops being flat.