September 2, 2026 · 7 min read · by Quanta Calculator

TDEE and Energy Balance: The Arithmetic of Maintenance Calories

How a TDEE estimate is actually built — Mifflin-St Jeor BMR, activity multipliers, the thermic effect of food — and how wide its error bars really are

Minimalist geometric illustration of stacked energy blocks, a small flame and a balance beam in warm amber tones

Total daily energy expenditure — TDEE — is the number of calories your body burns across a full 24 hours: running your organs, digesting your meals, powering your training, and covering every unplanned movement in between. It is also, by definition, your maintenance number: eat that many calories on average and body mass holds roughly steady over time, which is why "TDEE" and "maintenance calories" name the same quantity. Nobody measures it outside a laboratory. Every consumer tool, including the Quanta TDEE calculator, estimates it in two moves: predict resting expenditure from a regression equation, then scale it up with an activity multiplier.

That two-step construction matters more than the final figure, because each step carries its own uncertainty and the uncertainties compound. This guide assembles the estimate component by component — basal metabolic rate, the thermic effect of food, deliberate exercise, background movement — runs the actual arithmetic, and puts honest widths on every piece.

Four components, one number

Physiologists split a day's expenditure into four parts, and each part responds to different inputs:

Component	What it covers	Typical size
BMR — basal metabolic rate	organs and cellular upkeep at complete rest	roughly 60–70% of the total
TEF — thermic effect of food	digesting, absorbing, storing what you eat	roughly 8–12% of intake
EAT — exercise activity	deliberate training sessions	depends entirely on the training
NEAT — non-exercise activity	walking, standing, fidgeting, chores	the most variable component of all

BMR dominates the total, but NEAT is the wildcard: James Levine's research at the Mayo Clinic found it can differ by as much as roughly 2,000 kcal/day between adults of similar body size. An activity multiplier bundles EAT and NEAT into a single coefficient — convenient, and precisely where the estimate is softest.

Step one: the BMR floor

The foundation is the energy a body would burn lying still, awake, unfed, in a comfortable room — the round-the-clock cost of a heart, a brain, kidneys and a liver. Since almost nobody has this measured in a metabolic ward, it gets predicted instead. The standard equation is Mifflin-St Jeor, fitted in 1990 to indirect-calorimetry measurements of 498 healthy adults aged 19 to 78:

BMR = 10 × weight (kg) + 6.25 × height (cm) − 5 × age (years) + s

TDEE = BMR × activity factor, where s is +5 for males, −161 for females, and the factor runs 1.2 to 1.9

Take a 30-year-old male at 70 kg and 175 cm. The three terms are 10 × 70 = 700, then 6.25 × 175 = 1,093.75, then 5 × 30 = 150. Assembled: 700 + 1,093.75 − 150 + 5 = 1,648.75 kcal/day. A female with identical stats swaps +5 for −161; those constants sit 5 − (−161) = 166 kcal apart, so her prediction is 1,648.75 − 166 = 1,482.75 kcal/day. The BMR calculator displays these as 1,649 and 1,483 — whole numbers on purpose, because the equation's own error dwarfs any decimal. Note also what the age term does: every year subtracts 5 kcal/day from the prediction, or 5 × 10 = 50 kcal/day per decade, the equation's stand-in for the gradual loss of metabolically active lean tissue with age.

Step two: the multiplier, where the softness lives

The second step multiplies that floor by a factor describing how active a typical week is. Applied to the worked BMR of 1,648.75:

Activity band	Factor	Estimated TDEE (kcal/day)
Sedentary — desk job, little exercise	1.2	1,979
Lightly active — light exercise 1–3 days/week	1.375	2,267
Moderately active — exercise 3–5 days/week	1.55	2,556
Very active — hard exercise 6–7 days/week	1.725	2,844
Extra active — hard training plus a physical job	1.9	3,133

Each row is the same BMR multiplied out — 1,648.75 × 1.55 = 2,555.56, displayed as 2,556 — and each adjacent pair of rows sits 288–289 kcal apart (for instance 2,556 − 2,267 = 289). Selecting one band too high therefore inflates the estimate by nearly 300 kcal/day, and the full spread from sedentary to extra active is 3,133 − 1,979 = 1,154 kcal/day. This is the single most consequential input a person controls, and the tool's own guidance is blunt about the common failure: people who train hard a few days a week but sit the other 23 hours routinely pick a band above the one their week actually resembles.

Digestion's hidden line item

The thermic effect of food is the energy spent processing intake, typically 8–12% of it. At the moderately-active example's 2,556 kcal, that is somewhere between 2,556 × 0.08 = 204.48 and 2,556 × 0.12 = 306.72 kcal/day spent on digestion alone. It also depends on composition: processing protein costs roughly 20–30% of protein's own calories, carbohydrate 5–10%, and fat 0–3%. The multiplier framework never itemises any of this — TEF rides silently inside the activity factor — which is fine for estimation, but it means two diets of identical calories and different composition produce genuinely different expenditures that the equation cannot see.

How wide the error bars really are

The 2005 Frankenfield systematic review — the evidence analysis that made Mifflin-St Jeor the default clinical equation — found it lands within ±10% of measured resting expenditure in roughly 82% of non-obese and 70% of obese adults. That is the best record of the common equations, and it still means the prediction misses by more than 10% for roughly one non-obese adult in five (100 − 82 = 18).

Put widths on the worked example. Ten percent of the BMR is 1,648.75 × 0.10 = 164.875 kcal, so the honest form of that step is 1,649 ± 165: anywhere from about 1,484 to 1,814 kcal/day. Carry both edges through the same 1.55 factor and the TDEE range becomes (1,648.75 − 164.875) × 1.55 = 2,300.0 up to (1,648.75 + 164.875) × 1.55 = 2,811.1 — a band 511 kcal wide (2,811 − 2,300 = 511) before the multiplier choice is even questioned. The calculator's guidance accordingly frames its output as within ±10–15% of true maintenance for most people. The instrument that narrows the band is not a better formula; it is a scale read as a weekly average against a known, steady intake, since a few weeks of flat weight locates true maintenance more precisely than any regression can.

What "energy balance" adds

Energy balance is the signed difference between intake and expenditure, and the calorie deficit calculator makes the sign concrete using the field's oldest convention: Wishnofsky's 1958 estimate that a pound of adipose tissue stores roughly 3,500 kcal. A weekly rate of change converts to a daily gap as rate × 3,500 ÷ 7, so one pound per week corresponds to 3,500 ÷ 7 = 500 kcal/day and two pounds per week to 2 × 3,500 ÷ 7 = 1,000 kcal/day.

Two caveats keep that arithmetic honest. The rule is linear and bodies are not: Kevin Hall's 2011 dynamic modelling in the Lancet showed the 3,500-kcal rule overpredicts cumulative change by roughly 50% over a one-year horizon, because a smaller body spends less — and adaptive thermogenesis lowers resting expenditure a further 5–15% beyond what the lost mass alone predicts. Maintenance is a moving target, so any gap measured against it moves too. Second, the tool refuses to let arithmetic imply endorsement: it raises a safety flag whenever a plan exceeds 2 lb/week, implies a gap above 1,000 kcal/day, or pushes intake below its 1,500 kcal/day floor — thresholds drawn from ACSM and NIH clinical guidance, not from the formula.

A boundary worth stating plainly: everything above is estimation, not guidance. What a maintenance estimate or a deficit figure should mean for a particular person's eating — especially anyone with a medical condition, a history of disordered eating, or a clinician-managed weight plan — is a question for a registered dietitian or physician. The calculators compute; they do not prescribe, and neither does this post.

Read it as a range, not a number

A TDEE estimate is best understood as the centre of an interval: for the worked example, 2,556 kcal/day with realistic edges near 2,300 and 2,811. Held that way, the number does its real job — a starting hypothesis that a few weeks of steady intake and weekly-average weigh-ins will confirm, shift, or quietly overrule. Every coefficient in this guide also sits on the tool pages across Quanta, next to the studies it was fitted from, because an estimate is only as honest as the working it shows. Questions about the method, or about anything this piece left unclear, are welcome through the contact page.