Why TDEE calculators disagree

Cause one: the resting equation, worth about 157 kcal

For the reference body, Mifflin-St Jeor gives 1,780 kcal, the 1984 Roza and Shizgal revision of Harris-Benedict gives 1,854, and Cunningham's equation - the one textbooks call Katch-McArdle - gives 1,697 when fed a lean mass of 61.42 kg from Boer. The spread is 157 kcal, about 9 percent of the smallest figure. Note the conditional in that sentence: Katch-McArdle needs a lean body mass figure, and swapping Boer's estimate for Hume's 57.79 kg drops it to 1,618 and widens the spread to 236. An equation that depends on another estimate inherits that estimate's disagreement too.

Cause two: the activity multiplier, worth about 1,246 kcal

This is the big one, and it is the layer with no source. Running the same 1,780 kcal resting rate up the conventional ladder gives 2,136 at 1.2 and 3,382 at 1.9 - a range of 1,246 kcal a day, eight times the disagreement between the equations underneath. One rung is worth about 312 kcal for this body. Since nobody published the ladder, nobody published an error term for it either, and two calculators that describe the rungs slightly differently in their dropdown will send the same person to different rows in good faith.

Cause three: nobody shows the band, which is worth 552 kcal

Frankenfield 2005 assessed equations by whether they landed within 10 percent of a measured resting rate. Applying that to the reference body turns 1,780 into 1,602 to 1,958, and at the 1.55 multiplier that is a TDEE of 2,483 to 3,035 - a 552 kcal band, more than three times the spread between the three equations people argue about. Most calculators print a single figure to the calorie. The difference between two of those figures is usually smaller than the uncertainty either one is hiding.

What follows from this

A difference between two calculators is a difference of assumptions, not evidence that one is broken. If you want to compare, compare like for like: same equation, same multiplier, same units. If you want a figure you can act on, the estimate is a starting point and the scale is the instrument - a steady intake held for two or three weeks tells you more about your own expenditure than any equation can, because it measures the one body the equations were only averaging over.

Official sources

• Mifflin, St Jeor et al., A new predictive equation for resting energy expenditure, Am J Clin Nutr 1990
• Roza & Shizgal, The Harris Benedict equation reevaluated, Am J Clin Nutr 1984
• Cunningham, Body composition as a determinant of energy expenditure, Am J Clin Nutr 1991 — the equation usually called Katch-McArdle
• Frankenfield, Roth-Yousey & Compher, Comparison of predictive equations for resting metabolic rate, J Am Diet Assoc 2005
• FAO/WHO/UNU Expert Consultation, Human energy requirements (2004) — physical activity level bands
• Boer, Estimated lean body mass as an index for normalization of body fluid volumes, Am J Physiol 1984

Frequently asked questions