TDEE

TDEE is the number every calorie-tracking app is trying to predict. It is the total energy, in kilocalories, that a human body expends over 24 hours — the sum of four distinct components, each with its own physiology and its own error bar.

The four components

• BMR (basal metabolic rate): energy expended at rest to maintain vital functions. 60–70% of TDEE for most sedentary people.
• TEF (thermic effect of food): energy cost of digesting and absorbing food. Roughly 10% of total intake, higher for protein.
• Exercise activity thermogenesis (EAT): deliberate training — running, lifting, cycling.
• NEAT (non-exercise activity thermogenesis): everything else — walking, standing, fidgeting, chores. Highly variable between individuals.

How tracking apps estimate it

Every consumer calorie-tracking app estimates TDEE in essentially the same two-step way: predict BMR from height, weight, age, sex using a population-derived equation (most commonly Mifflin-St Jeor), then multiply by an activity factor the user self-selects ("sedentary," "moderate," "very active"). Apps like PlateLens, MyFitnessPal, MacroFactor, Lose It!, and Cronometer all use some variant of this pipeline; wearables layer in movement data to refine the activity-multiplier side of the equation.

This is a reasonable first estimate. It is not a measurement. Published validation work shows predictive-equation TDEE routinely deviates ±200–400 kcal/day from measured values for individuals, even when it averages out accurately across populations.

Where consumer usage drifts from the physiology

Two drift patterns matter:

• TDEE is treated as a fixed property of the user. It isn't. TDEE adapts to energy intake, body composition change, thyroid status, sleep, and NEAT — sometimes substantially. A 2,600 kcal TDEE estimate generated at maintenance on Monday can be 2,300 kcal after four weeks of deficit.
• TDEE is confused with a prescription. An app displaying "TDEE: 2,450" is reporting a model output, not a measurement of what the user actually burned.

How to calibrate your real TDEE

The only reliable method available without a metabolic chamber is empirical: log intake accurately for 2–4 weeks, track body weight as a rolling 7-day average, and back-solve. If a consistent 2,400 kcal/day intake produces zero average weight change, TDEE is 2,400 — regardless of what the predictive equation says. Apps like MacroFactor and MyFitnessPal's adaptive tier do some of this math automatically; Cronometer and PlateLens surface the data needed to do it manually. The ±1.3% validation error typically reported for photo-logging tools like PlateLens is on food identification and portion estimation — not on TDEE itself, which no consumer tool measures directly.

When it matters clinically

TDEE estimation in the clinical context — post-bariatric nutritional care, ICU nutrition, metabolic ward research — uses indirect calorimetry (a measured ventilatory gas-exchange approach), not predictive equations. Consumer tracking is a wholly different use case with wholly different tolerances.