Development and validation of new predictive equations for the resting metabolic rate of older adults aged ≥65 y

Judi Porter, Leigh C. Ward, Kay Nguo, Zoe Davidson, Simone Gibson, Ross Prentice, Marian L. Neuhouser, Helen Truby

Research output: Contribution to journalArticleResearchpeer-review

2 Citations (Scopus)


Background: The aging process alters the resting metabolic rate (RMR), but it still accounts for 50%–70% of the total energy needs. The rising proportion of older adults, especially those over 80 y of age, underpins the need for a simple, rapid method to estimate the energy needs of older adults. Objectives: This research aimed to generate and validate new RMR equations specifically for older adults and to report their performance and accuracy. Methods: Data were sourced to form an international dataset of adults aged ≥65 y (n = 1686, 38.5% male) where RMR was measured using the reference method of indirect calorimetry. Multiple regression was used to predict RMR from age (y), sex, weight (kg), and height (cm). Double cross-validation in a randomized, sex-stratified, age-matched 50:50 split and leave one out cross-validation were performed. The newly generated prediction equations were compared with the existing commonly used equations. Results: The new prediction equation for males and females aged ≥65 y had an overall improved performance, albeit marginally, when compared with the existing equations. It is described as follows: RMR (kJ/d) = 31.524 × W (kg) + 25.851 × H (cm) − 24.432 × Age (y) + 486.268 × Sex (M = 1, F = 0) + 530.557. Equations stratified by age (65–79.9 y and >80 y) and sex are also provided. The newly created equation estimates RMR within a population mean prediction bias of ∼50 kJ/d (∼1%) for those aged ≥65 y. Accuracy was reduced in adults aged ≥80 y (∼100 kJ/d, ∼2%) but was still within the clinically acceptable range for both males and females. Limits of agreement indicated a poorer performance at an individual level with 1.96-SD limits of approximately ±25%. Conclusions: The new equations, using simple measures of weight, height, and age, improved the accuracy in the prediction of RMR in populations in clinical practice. However, no equation performs optimally at the individual level.

Original languageEnglish
Pages (from-to)1164-1173
Number of pages10
JournalThe American Journal of Clinical Nutrition
Issue number6
Publication statusPublished - Jun 2023


  • anthropometric measures
  • older adults
  • prediction
  • resting metabolic rate
  • RMR

Cite this