Yes — calories in, calories out is true, in the way that a bank balance is true: what comes in minus what goes out equals the change in what is stored. That is conservation of energy and no diet can break it. What is not true is the version most people are sold, in which "calories out" is a fixed number you read off a calculator once and plan against for months. The "out" side moves — with your weight, your activity and, awkwardly, with the deficit itself.
The equation
calories in − calories out = change in stored energy
Three terms. "In" is what you eat and absorb. "Out" is total daily energy expenditure — resting metabolism, digestion, training and all other movement. "Change in stored energy" is what the body added to or drew from its reserves, which over weeks shows up as a change in body weight.
A few honest notes on each term:
| Term | What it really is | How well you can know it |
|---|---|---|
| Calories in | energy absorbed from food | logs are typically off by a tenth or more; labels are rounded |
| Calories out | BMR + TEF + exercise + NEAT | a calculator estimate is 100–200 kcal off, sometimes 400–500 |
| Change in stored energy | fat + lean tissue + glycogen + water, each with a different energy value | scale weight is noisy; trend weight over weeks is usable |
The arithmetic is exact. The inputs are not. Every disagreement about "CICO" is really a disagreement about how well the terms can be measured, and about whether "out" stays put.
Where "out" hides adaptation
It does not stay put. Three mechanisms, each with a measured basis:
- NEAT drops when you underfeed. Non-exercise activity thermogenesis — fidgeting, standing, walking about — is modulated by energy balance: it increases with overfeeding and decreases with underfeeding1. Someone in a deficit moves less without deciding to. The calculator estimate made in week one assumed the week-one NEAT.
- A lighter body costs less. Every kilogram lost is 10 kcal a day less of resting expenditure in the Mifflin-St Jeor equation, and less energy to carry up the stairs. Ten kilograms down is roughly 100 kcal a day off BMR before any multiplier, and more off the total.
- Total expenditure is constrained, not additive. In 332 adults across five populations, total energy expenditure measured with doubly labelled water rose with physical activity at the low end and then plateaued at the upper range of activity2. Adding more movement did not keep adding expenditure indefinitely; the body adapted to keep the total within a narrower band than an additive model predicts.
None of these is a metabolism that shuts down and makes a deficit impossible. Energy balance still holds. It is just that the "out" number quietly shrinks, so a fixed intake becomes a smaller deficit over time — and the arithmetic then correctly predicts slower change.
In plain terms: the equation never lies, but one of its terms keeps changing behind your back.
Why 7,700 kcal per kg is a planning number, not a law
Our calculator turns a weekly rate into a daily deficit with daily deficit = weekly kg × 7,700 ÷ 7 — for an 85 kg person at the moderate rate, 85 × 0.005 = 0.425 kg a week, 0.425 × 7,700 ÷ 7 ≈ 468 kcal a day. The 7,700 (about 3,500 kcal per pound) is the energy in a kilogram of adipose tissue, which is mostly fat but not entirely.
Real weight change is not pure adipose tissue:
| Tissue | Approximate energy per kg | When it dominates a weight change |
|---|---|---|
| Body fat (adipose) | ≈ 9,400 kcal | slow loss or gain in someone with fat to lose |
| Lean tissue (mostly water and protein) | ≈ 1,800 kcal | fast loss, very lean people, hard training |
| Glycogen with its water; gut contents; salt water | ≈ 0 kcal per kg of scale change | the first days of a diet; the morning after a big meal |
These are modelling figures, not measurements of you. The consequence: the same 468 kcal daily deficit "buys" 0.35 kg a week if it all comes from adipose tissue and far more scale change if some comes from lean tissue and water — which is why the first week of a diet often looks dramatic and the fourth looks like nothing. Adaptive expenditure models deal with this by assuming a fat fraction that depends on the rate of change (see how apps estimate energy expenditure). A calculator cannot; it uses 7,700 as a planning constant, and the honest description of that constant is "close enough to set a starting target, not accurate enough to predict the scale".
What it means for tracking
The practical reading of all this:
- Use the equation backwards. Forwards — predicting weight from a calculator's "out" — it is only as good as the calculator. Backwards — reading your real "out" from logged intake and trend weight — it is a measurement of you. That is exactly what MacroDiet's expenditure estimate does once it has three consecutive days of logging and keeps doing with 3 food days and 1 weigh-in per week.
- Expect the target to need revising. Because "out" falls during a deficit, the intake that produced 0.4 kg a week in month one produces less in month three. That is the equation working, not failing. Our rates are set as a percentage of body weight for the same reason — the absolute deficit shrinks as you do.
- Judge by the trend, not the day. Water and glycogen can move the scale by a kilogram overnight, which is 7,700 kcal of "energy" that never existed. Three or four weeks of trend weight is the shortest window in which the equation's stored-energy term is legible.
This is arithmetic and mechanism, not medical advice. The energy values of tissue are model constants; the rate at which anyone's expenditure adapts is individual and not something a page can tell you.
The short version
Calories in, calories out is conservation of energy — always true — but "out" is not a fixed number. It falls as NEAT drops, as the body gets lighter, and because total expenditure plateaus rather than growing with every extra hour of activity. 7,700 kcal per kg is a planning constant for body fat, not a promise about the scale. Use the equation backwards: logged intake plus trend weight tells you your real expenditure, and that number is allowed to change. Educational overview only — not medical advice.