MET estimates improve when scaled to you

Multiplying Compendium METs by body weight and time gives a workable estimate of a workout's energy cost. The standard resting value behind METs overstated measured resting energy use by about 20 to 35% on average, and errors were largest in low-fit, overweight and older people. Scaling METs to a person's predicted resting metabolic rate brought estimates closer to measured values.

Why it works

A MET expresses energy use as a multiple of a standard resting value (3.5 mL of oxygen per kg per minute, or 1 kcal per kg per hour). When a person's true resting rate is lower than that standard, as it was on average in Byrne 2005 and Kozey 2010, estimates built on the standard value misstate their energy use. Body composition explained most of the variation in resting oxygen use (Byrne 2005).

What it means in practice

Use Compendium METs times weight times duration as the workout estimate, and scale MET values by the user's predicted resting metabolic rate where possible. Present calorie numbers as estimates, not measurements.

The evidence

• Byrne, N.M., Hills, A.P., Hunter, G.R. et al. (2005). Metabolic equivalent: one size does not fit all. Journal of Applied Physiology.

  Measured resting values averaged 2.6 ml O2/kg/min and 0.84 kcal/kg/h, significantly lower than the standard 1-MET values. In this large heterogeneous sample the 3.5 ml/kg/min convention overestimated resting oxygen consumption by 35% on average, and the 1 kcal/kg/h convention overestimated resting energy expenditure by 20%. Body composition explained 62% of the variance in resting oxygen consumption versus 14% for age. Using measured or predicted resting metabolic rate as a correction factor adjusted for individual differences when estimating walking energy cost.

  n=769

• Kozey, S., Lyden, K., Staudenmayer, J. et al. (2010). Errors in MET estimates of physical activities using 3.5 ml x kg(-1) x min(-1) as the baseline oxygen consumption. Journal of Physical Activity and Health.

  Measured METs averaged 6.1. Standard METs based on 3.5 ml/kg/min were 0.51 METs lower than measured values, while corrected METs did not differ from measured. Standard METs misclassified 12.2% of activities by intensity versus 8.6% for METs based on predicted RMR. Errors were largest for low-fit, overweight and older individuals, and the authors concluded these errors disproportionately affect the least active groups.

  n=252

• Ainsworth, B.E., Haskell, W.L., Herrmann, S.D. et al. (2011). 2011 Compendium of Physical Activities: A Second Update of Codes and MET Values. Medicine and Science in Sports and Exercise.

  This is a reference update, not an experiment. It lists the energy cost (in METs) of specific physical activities so studies can compare self-reported activity on a common scale. The 2011 version contains 821 activity codes, including 217 new ones, and 68% (561 of 821) carry directly measured MET values rather than estimates. About half of the codes from the 2000 edition were revised to sharpen definitions or update values, with updated MET values accounting for 73% of those changes.

• Herrmann, S.D., Willis, E.A., Ainsworth, B.E. et al. (2024). 2024 Adult Compendium of Physical Activities: A Third Update of the Energy Costs of Human Activities. Journal of Sport and Health Science.

  This is a reference update, not an experiment. It catalogs the energy cost (in METs, a multiple of resting metabolism) of 1,114 physical activities across categories like leisure, sport, transportation, home, and work. The authors screened 32,173 abstracts and 1,507 full papers, then pulled 2,356 measured energy-expenditure values from 701 studies to build it. Compared with the prior version it adds 303 new activity codes and modifies 176 existing ones, giving more entries backed by directly measured (rather than estimated) values.

Why we call confidence medium

The Compendium of Physical Activities (Ainsworth 2011, Herrmann 2024) is the standard catalogue of MET values, and the 2024 edition draws on measured values from 701 studies, but it does not test individual accuracy. In 769 adults (Byrne 2005), measured resting values averaged 2.6 mL O2/kg/min, so the 3.5 mL/kg/min convention overestimated resting oxygen use by 35% and the 1 kcal/kg/h convention overestimated resting energy expenditure by 20%. Using measured or predicted resting metabolic rate as a correction factor adjusted for individual differences. In 252 adults across 2,555 activities (Kozey 2010), standard METs were 0.51 MET lower than measured values and misclassified 12.2% of activities by intensity, versus 8.6% for METs based on predicted resting metabolic rate, with the largest errors in low-fit, overweight and older people. Medium: the method is standard and its main bias is known and correctable, but no study validates MET times weight times duration against measured workout energy for individual runners.

Where it applies

Healthy adults. The 2024 Compendium covers ages 19 to 59, Byrne 2005 studied adults 18 to 74 (mostly women, some obese), and Kozey 2010 adults in their 20s to 50s.