What a TDEE number actually tells you
A calculator says 2,410 calories a day with reassuring precision. Here's what that figure is made of, how wide its true error bars are, and why the fourth digit is fiction.
Total Daily Energy Expenditure is an estimate of how many calories you burn in a day. Every calculator that produces one is doing the same two-step: estimate your resting metabolism from body data, then multiply by an activity factor. Understanding both halves tells you how much to trust the answer.
Step one: basal metabolic rate
BMR is what you'd burn lying still all day — keeping a body at temperature, running a brain, circulating blood. For most people it's 60–70% of total expenditure, so it dominates the result.
The two equations in common use:
Mifflin-St Jeor — the default in most calculators, and the better-validated one for the general population:
- Men: (10 × kg) + (6.25 × cm) − (5 × age) + 5
- Women: (10 × kg) + (6.25 × cm) − (5 × age) − 161
Katch-McArdle — uses lean body mass instead of total weight, so it needs a body-fat percentage:
- 370 + (21.6 × lean mass in kg)
Katch-McArdle is more accurate if your body-fat figure is accurate, because muscle and fat have very different metabolic rates. That's a real if: bathroom scale bioimpedance readings swing several percentage points with hydration alone. With a guessed body-fat number, Mifflin-St Jeor is the safer bet.
The honest error bar: Mifflin-St Jeor predicts measured BMR within 10% for roughly 80% of people. So a "1,700 calorie" BMR is really somewhere around 1,530–1,870 — and for the remaining one in five, further out than that.
Step two: the activity multiplier
The standard ladder, unchanged since the 1990s:
| Level | × | Roughly |
|---|---|---|
| Sedentary | 1.2 | Desk job, little deliberate exercise |
| Lightly active | 1.375 | Light exercise 1–3 days a week |
| Moderately active | 1.55 | Moderate exercise 3–5 days |
| Very active | 1.725 | Hard exercise 6–7 days |
| Extra active | 1.9 | Physical job plus training |
Notice the gap between rungs: about 175 calories at a 1,700 BMR. Sitting between two descriptions — as most people genuinely do — puts a couple of hundred calories of uncertainty into the answer before anything else goes wrong. And the usual failure is optimism: three gym sessions a week does not make a desk worker "moderately active" if the other 165 hours are spent sitting.
The two components nobody accounts for
TEF — the cost of digestion
Digesting food burns energy, about 10% of intake on a mixed diet. Protein is expensive to process (20–30% of its calories), carbohydrate middling (5–10%), fat cheap (0–3%). The activity multipliers roll this in implicitly, which is fine — but it does mean a very high-protein diet runs a slightly higher expenditure than the same calories from fat.
NEAT — everything that isn't exercise
Non-Exercise Activity Thermogenesis: walking to the printer, fidgeting, standing, carrying shopping, gesturing while talking. This is the most variable component in the whole picture. Studies of overfeeding have found NEAT differences of up to 2,000 calories a day between individuals — dwarfing any gym session.
NEAT also adapts. Eat less for a few weeks and NEAT quietly falls: you move less, unconsciously. That's a large part of why weight loss plateaus while intake stays honest, and why your TDEE today isn't your TDEE after three months of dieting.
How to actually use the number
Treat the calculator output as a starting hypothesis, not a measurement. The procedure that works:
- Calculate a TDEE estimate.
- Eat at that level for two to three weeks, tracking honestly.
- Weigh yourself daily, same conditions, and compare weekly averages — daily weight swings 1–2 kg on water and gut contents alone, which is far larger than any real change over a day.
- Weight stable → the estimate was about right, and now you have a measured maintenance figure. Drifting → adjust by 100–200 calories and repeat.
Two or three weeks of your own data beats any equation, because it measures you rather than a population average that happens to share your height.
Why the calorie numbers on both sides are soft
Expenditure is an estimate — but so is intake. Food labels are permitted a tolerance of around 20% in most jurisdictions. Portion estimates by eye are routinely out by 20–30%. The calorie readout on cardio machines is essentially decorative, typically overstating by a third.
So you are comparing an estimate with an error bar against another estimate with an error bar. This is not a reason to give up on tracking — consistency still reveals the trend, and the trend is what you act on. It is a reason to stop treating a 50-calorie difference as meaningful, and to distrust any tool that reports your requirement to four significant figures.
These are population-average equations, not medical advice. Pregnancy, thyroid conditions, medication and a history of dieting all shift the picture in ways no calculator models. For anything medical, talk to someone qualified who can look at you rather than at your height.