Maintenance Calories After Weight Loss

Your new maintenance is a smaller body, plus an argument about the remainder#

After a diet, the intake that holds your weight steady is lower than the one that held it steady before — and the large, uncontroversial reason is that there is less of you to run and less of you to carry around. The disputed part is the remainder: whether a weight-reduced body burns less than its new size predicts, permanently, as a kind of penalty for having dieted. That question decides whether you set your new number by arithmetic or by superstition, and it has a better answer than the internet suggests.

The short version is that the remainder is real during the diet, contested afterwards, and small enough that no practical plan should be built around it. What should be built around instead is a measurement — a stepped ramp out of the deficit that takes four to six weeks and gives you a number denominated in your own life rather than in a formula's population average. This article is that ramp, the evidence for how low the number really sits, and what the maintenance trials say about what to put in it.

The best-matched comparison could not find the penalty#

Start with the study that should be far better known than it is, because it was designed specifically to test the folklore. Twenty-four overweight postmenopausal women were studied across four 10-day phases: before weight loss, and again after losing a mean of 12.9 ± 2.0 kg and returning to energy balance — that is, once they had stopped actively losing. Their resting metabolic rate was then compared against 24 never-overweight control subjects matched for the body they now had.

The weight-reduced women came in at 4,771 ± 414 kJ/day. The women who had never been overweight came in at 4,955 ± 414 kJ/day, P = 0.14¹. That gap works out at about 44 kcal a day — my conversion of their two figures, and a difference the study could not tell apart from zero. The follow-up is the part that matters for planning: within this cohort, having a lower resting rate carried no signal about who would regain most across the following four years (r = 0.27, ns). Whatever makes maintenance difficult, the metabolic penalty did not identify the people it happened to.

The famous counter-case is real and belongs on the page: fourteen Biggest Loser contestants who lost 58.3 kg in 30 weeks were still running about 499 ± 207 kcal/day below prediction six years later⁵. Twenty-four postmenopausal women losing 13 kg and returning to weight stability, versus fourteen people who lost the mass of a child under television conditions, are not measuring the same phenomenon — the size of the intervention and the energy-balance state at the moment of measurement are what separate them, and that argument is worked through in the starvation-mode myth. For someone finishing an ordinary diet, Weinsier's population is the closer match.

What actually makes maintenance hard, and it is not the burn#

The reason "my metabolism is broken" survives despite results like Weinsier's is that maintenance genuinely is hard, so people reach for the only mechanism they have heard of. The larger mechanism sits on the other side of the ledger and cannot be fixed by any number you set.

When a deficit is opened covertly — by a drug that dumps glucose into urine, in 153 patients who were not dieting — the modelled rise in food intake came out at roughly 100 kcal a day per kilogram shed, which the authors put at over triple the matching change in expenditure². That ratio, its caveats and its provenance are the subject of how metabolic adaptation causes plateaus; the expenditure half is costed in why your calorie needs drop as you lose weight.

What it means here is narrow and worth stating. The recalculation you are about to do addresses the smaller of the two forces acting on your new weight. It is still the one worth doing, because it is the one that takes instructions.

The ramp: four to six weeks, not one adjustment#

The common failure is jumping from your deficit intake to a calculated maintenance figure in a single step, then reading the following week's scale as a verdict. That week contains a glycogen refill and its bound water, so it will show a gain of a kilogram or more regardless of what you did, and it tells you nothing about energy balance.

A stepped ramp separates the two signals:

Week	What to do	What the scale will show	What it means
1–2	Add ~100–150 kcal/day to your deficit intake	+0.5–1.5 kg in the first few days, then flat	Glycogen and water. Ignore the level; watch the slope.
3–4	Add another ~100–150 kcal/day	Slope near zero, or still gently down	You are approaching the number from below
5–6	Hold; add one more step only if the trend is still falling	Flat 7-day averages, two weeks running	This intake is your maintenance, in your logging units
After	Re-run the loop when your weight or activity changes materially	—	The figure has a shelf life

Two rules make the ramp readable. Judge on 7-day averages, never on mornings, and discard the first three to five days after any change while fluid settles. And keep everything else fixed while you do it: the same logging habits, the same training, the same protein. The number you land on is calibrated to your instrument — if you switch apps or start weighing what you used to eyeball, you have changed the units and owe yourself another loop. The general version of that back-calculation, for people not coming off a diet, is in finding your maintenance calories.

Approaching from below matters more than it sounds. Coming up in steps, the worst case is that you spend an extra fortnight in a small deficit. Coming down from an overshoot, the worst case is several kilograms back and the sense that maintenance "didn't work" — and the appetite figure above says the overshoot is the more likely direction of error.

What to put in the number: 26 weeks says one thing, 148 weeks another#

Composition at maintenance has been tested properly, twice, and the two results are usually presented as a contradiction. They are not.

In the DiOGenes trial, 773 adults who had already lost a mean of 11.0 kg on an 800-kcal diet were randomized to five ad libitum maintenance diets for 26 weeks in a protein × glycemic-index factorial. The high-protein arms regained 0.93 kg less than the low-protein arms (95% CI 0.31–1.55, P = 0.003) and the low-glycemic-index arms 0.95 kg less than the high-GI arms (95% CI 0.33–1.57, P = 0.003). Only the low-protein/high-GI combination produced significant regain at all. And the largest effect was on staying in the study: 26.4% and 25.6% dropout in the high-protein and low-GI groups against 37.4% in the low-protein/high-GI group³.

Then PREVIEW ran a similar contrast — high-protein/low-GI against moderate-protein/moderate-GI — through a 148-week maintenance phase in 2,223 adults who had lost at least 8%. Weight regain did not differ between the diet groups (P = 0.630). What did differ was hunger, from week 52 onward, in favour of high-protein/low-GI⁴. Several authors on that paper disclose glycemic-index research positions, food-industry honoraria or advisory-board seats, which is worth knowing about a result that happens to favour the null.

Six months bought about a kilogram. Three years bought less hunger and no kilograms. Both trials are describing an adherence effect that runs out of runway.

Read them together and the practical instruction is the same either way: higher protein and lower-glycemic-index carbohydrate at maintenance make the diet easier to keep running, which is worth roughly a kilogram over half a year and shows up as reduced hunger rather than reduced weight over three. That is a modest, useful claim, and it is not the metabolic advantage it is usually sold as.

Treat the answer as a band with a trip line#

One last framing, because the ramp produces a number and a number invites false precision. Your logging carries error, your daily weight carries noise, and your maintenance genuinely drifts with season, activity and age. What the ramp actually locates is a band — a range of a few hundred calories inside which your weight is stable enough that you cannot tell the difference on a monthly scale.

So define maintenance by a weight band rather than a calorie point: pick your settled weight, add a couple of kilograms, and treat crossing that upper line for a full week as the signal to tighten, not as a failure. That converts an unwinnable precision problem into a decision rule you can actually run, and it is the operating half of keeping the weight off. The stepped ramp is also the closest thing to what people mean by reverse dieting that survives contact with the evidence: a controlled approach to maintenance from below, with no claim that it raises your metabolism on the way.

FAQ#

How long after a diet should I wait before trusting my maintenance number?#

Four to six weeks of stepped increases, with the first three to five days after every change discarded. Body weight moves a kilogram or more on glycogen and its bound water when you add carbohydrate back, so the level immediately after a change is uninformative — only the slope of consecutive 7-day averages tells you whether you are still in a deficit. Two consecutive flat weekly averages at a fixed intake is the earliest point the number means anything.

Should I raise calories all at once or gradually after a diet?#

Gradually, in steps of roughly 100–150 kcal a day, because the errors are asymmetric. Approaching maintenance from below costs you at most a few extra weeks in a mild deficit; overshooting it costs you weight you then have to re-lose, and appetite is pushing in that direction — roughly 100 kcal/day of extra intake pressure for every kilogram you have lost².

Does eating more protein after a diet prevent regain?#

It helps modestly and mostly by making the plan easier to keep. Over 26 weeks, high-protein arms regained 0.93 kg less than low-protein arms and were far likelier to complete the study — 26.4% versus 37.4% dropout³. Over three years, the same contrast produced significantly lower hunger but no difference in weight regain at all⁴.

Sources#

This article was researched and drafted with AI assistance and reviewed for accuracy by the BurnWeek team. It is general information, not medical advice. How we research and correct our articles →