What is a calorie deficit, and how much per week is safe?

What a calorie deficit actually is

Your body spends energy all day, and most of that spending happens while you do nothing at all. Breathing, circulation, kidney filtration, the constant repair and replacement of cells — in a sedentary adult this baseline accounts for something like 60–70% of the daily total. Movement sits on top of it. So does digestion, which has a bill of its own: breaking food down costs energy, and the cost varies by nutrient, with 20–30% of the calories in protein lost as heat against roughly 0–3% for fat.

Eat less than that total and your body covers the shortfall from storage. That shortfall is the deficit. There is no second mechanism.

A deficit is an absolute number, not a percentage. Someone burning 2,400 kcal who eats 2,000 has built a 400 kcal deficit. The same 400 kcal is a far harsher cut for someone burning 1,800 — a sixth of the budget in one case, close to a quarter in the other. This is exactly where flat prescriptions like “1,200 calories for everyone” fall apart: they do nothing for a large man and push a small woman somewhere nobody should stay for long.

Where 7,700 comes from, and where it drifts

Fat tissue is not pure fat. Around 85–87% of it is triglyceride; the rest is water, connective tissue and cell structure. Since pure fat carries 9 kcal per gram, a kilogram of fat tissue works out to roughly 7,700 kcal — about 3,500 kcal per pound. Every weekly-target table is that number run in reverse:

• 500 kcal/day → 3,500 kcal/week → about 0.45 kg (1 lb)
• 750 kcal/day → 5,250 kcal/week → about 0.68 kg (1.5 lb)
• 1,100 kcal/day → 7,700 kcal/week → about 1 kg (2.2 lb)

The table starts out honest and drifts. Two reasons. Not everything you lose is fat — the harsher the deficit, the larger the share taken from muscle and water. And your expenditure does not hold still: a body that has dropped from 80 kg to 75 kg is a lighter body to carry, so the same walk now costs fewer calories.

Expect the rule to hold up reasonably well over six weeks and to overpromise badly over six months. It is a starting point. Not a schedule.

The margin hiding inside your maintenance number

You do not know your daily expenditure. You estimate it. FitDex estimates it with the Mifflin-St Jeor equation, because the 1919 Harris-Benedict equation runs systematically high against modern body compositions.

Mifflin-St Jeor is not exact either. It is expected to land within about ±10% of measured expenditure, which means an estimate of 2,400 kcal describes a real value somewhere between 2,160 and 2,640. Sit at the low end of that band, trust the table, cut 500 — and your actual deficit might be 260. When the scale refuses to move, this band deserves suspicion long before your willpower does.

There is a second, larger error and it runs the other way. Studies using doubly labelled water, the reference method for measuring real-world expenditure, have repeatedly found that people under-report what they eat by 20–30%, and that the gap widens as body weight rises. Nobody in those studies was lying. Portions drift, cooking oil goes uncounted, and the handful of nuts never makes it into the log.

The fix is not a better formula. Log honestly for two weeks, track the weight change, and back out your real maintenance from what actually happened — an estimate built from your own data beats any equation, because it has stopped being an estimate.

Choosing a weekly rate

The commonly recommended range is 0.5% to 1% of body weight per week. For an 80 kg person that is 0.4–0.8 kg; for a 60 kg person, 0.3–0.6 kg. Handing both of them the same absolute target asks proportionally twice as much from the lighter one.

Body fat percentage belongs in this decision too. Someone carrying a large fat store tolerates an aggressive deficit reasonably well, while someone already lean loses more muscle at the same rate, simply because there is less fat available to draw on. Leaner means gentler.

The bottom of the range looks slow on paper. Over twelve weeks, 0.4 kg a week still adds up to roughly 5 kg, and that 5 kg tends to stay lost longer than the 8 kg someone else squeezed out of the same twelve weeks, because a slower plan gives habits time to catch up with the scale.

The floor underneath the deficit

You cannot enlarge the deficit indefinitely, and the reason is arithmetic rather than moral. Below a certain intake, fitting your protein, fibre, iron and calcium requirements onto the plate becomes genuinely difficult; the plate shrinks but the requirements do not.

The second limit is quieter. Under a hard deficit, spontaneous movement falls away — fewer steps, less standing, less fidgeting. This line item is called NEAT, and the spread between individuals runs to several hundred calories a day. Push the deficit on paper to 800 while unknowingly dropping expenditure by 300 and what you are left with is 500.

FitDex applies a floor for these two reasons and warns when your target would fall below it. If you believe you need to go under that floor, the question has stopped being one for a calculator. It is one for a doctor.

Why the scale swings every morning

Most day-to-day variation on the scale is water, not fat. Every gram of glycogen stored in muscle and liver holds roughly 3 grams of water alongside it, so the morning after a high-carbohydrate day the scale can read 1–2 kg heavier with no change in body fat whatsoever. Salt, menstrual cycle, gut contents, inflammatory water retention after a hard session — all of it lands in the same column.

A 500 kcal daily deficit, meanwhile, is worth about 65 grams a day. The signal you are looking for is roughly a tenth the size of the noise it is swimming in.

Weighing more often does not fix this; changing the window does. Weigh under identical conditions — morning, after the bathroom, before eating — and compare weekly averages instead of individual days. If the two-week average has not moved, the deficit was never really there.

Where to start

Estimate your expenditure, pick a weekly rate, then find the daily intake that rate implies. All three are linked, which makes doing them separately an easy way to introduce an error; the tools below run the same formulas in sequence and show the results together.