TDEE for athletes: why the formula is only a starting point

Total daily energy expenditure (TDEE) is the energy a person burns in a day: resting metabolism, digestion, daily movement and training. Calculators estimate it from a resting metabolic rate (RMR) equation and an activity multiplier. For athletes this produces a usable first guess and little more. An adaptive TDEE, back-calculated from logged intake and the fasted weight trend, replaces the guess with a measurement after two to four weeks.

How wrong the formulas are

Jagim and colleagues (2018) measured RMR by indirect calorimetry in 50 college athletes and compared it with five prediction equations. All five underestimated measured RMR. The closest, the Cunningham equation, was off by about 165 kcal a day on average, and individual errors were larger: the root-mean-square error was 284 kcal for the best equation in men.

A 2023 systematic review and meta-analysis (O’Neill et al.) pooled 29 studies with 1,430 athletes and 100 equations. Some equations, including Mifflin-St Jeor, the default in many consumer calculators, significantly under- or overestimated RMR in athletes. Even the most precise equation (Ten Haaf) placed about 80% of athletes within 10% of their measured value, and most equations managed only 40 to 64%.

That is the resting part. The activity multiplier is a coarser guess still, and for a hybrid athlete it changes from a recovery week to a build week. Combined, a formula TDEE can be several hundred kcal a day away from reality, which is the size of an entire weight-loss deficit.

The adaptive method

Energy balance can be solved backwards. If intake is known and the weight trend is known, expenditure follows:

TDEE ≈ average daily intake + (trend weight loss per day × 7,700 kcal per kg) (the loss is negative when the trend rises)

An athlete eating 2,600 kcal a day whose trend falls 0.25 kg (0.55 lb) a week is losing about 0.036 kg a day, worth about 275 kcal. Their expenditure is about 2,875 kcal. The 7,700 kcal per kg (3,500 kcal per lb) figure is a simplification that holds over weeks better than over months (Hall, 2008), which is one more reason to keep recalculating.

• Data needed. Consistent food logging and fasted morning weigh-ins under fixed conditions (see water weight for athletes). Missing or partial days bias the result.
• Window. Two weeks gives a first estimate. Four weeks is noticeably more stable. Shorter windows are dominated by water.
• Blend, do not jump. A sensible approach starts from the formula and shifts weight towards the observed value as data accumulates, so one salty weekend cannot move the estimate by 500 kcal.
• Exclude unusual readings. Non-fasted weigh-ins, illness and travel days add noise without information.

Adding training without double counting

An adaptive TDEE already contains the average training of the window it was measured over. Adding every logged session on top counts that training twice. Two approaches avoid this.

1. Baseline plus training. Measure a baseline that excludes exercise, then add each day’s training burn. Smoothing that burn across the day of the session and the following one or two days matches when refuelling actually happens.
2. Average only. Use the adaptive TDEE as is and accept that hard weeks run a larger deficit and easy weeks a smaller one. This is simpler and works when training volume is steady.

For athletes who are cutting, the deficit itself can follow training load: smaller when accumulated fatigue is high, larger in recovery weeks. The reasoning is covered in losing weight while training.

Why TDEE drifts

Expenditure is not fixed during a diet. A lighter body burns less, non-exercise activity tends to fall, and adaptive thermogenesis lowers expenditure beyond what tissue loss explains. Training volume changes on top of that. A TDEE calculated once at the start of a cut is often the reason a plateau appears a few weeks later (see plateau or water?). A weekly recalculation from the latest four weeks of data keeps the target honest.

Tools

Any spreadsheet can run the adaptive formula. Some nutrition apps calculate it automatically. For a formula starting point, the TDEE and macro planner on aitrainingplan.app is free. A worked description of an adaptive, training-adjusted TDEE is on its adaptive TDEE page. The editor publishes his own daily predictions and the TDEE each one used in a public N=1 log.

FAQ

How accurate are TDEE calculators for athletes?

Not very. Resting metabolic rate equations tend to underestimate measured values in athletes, by about 165 kcal a day on average for the best equation in one study (Jagim et al., 2018), and activity multipliers add a much larger error on top. A formula is a reasonable first guess, off by several hundred kcal a day for many athletes.

What is an adaptive TDEE?

An estimate of total daily energy expenditure calculated from what an athlete actually eats and how their weight trend moves. Average daily intake plus the weight lost on the trend, converted to energy (about 7,700 kcal per kg, 3,500 kcal per lb) gives the expenditure that must have happened. It needs two to four weeks of consistent food logging and fasted weigh-ins.

Should TDEE change on training days?

The baseline should not jump day to day, because the scale cannot resolve single days. Training burn is better handled separately: added on top of a baseline and smoothed over two or three days, since refuelling a long session happens on the day and the days after it.

How often should TDEE be recalculated?

Weekly is enough. Expenditure drifts as body weight falls, as training volume changes and through adaptive thermogenesis, so a TDEE set at the start of a diet is usually wrong a few kilograms later.