Work

Comparison of Resting Energy Expenditure Prediction Equations and Indirect Calorimetry Among Adults with Severe Obesity

Year: 2025

Type: article

Abstract: Background Proper estimations of resting energy expenditure (REE) are important for developing weight management strategies, but it is unclear which prediction equations best estimate REE for those with severe obesity. Objective This validation study tested 11 previously validated REE prediction equations to determine which equations estimate REE with the least bias and most precision in participants with severe obesity. Methods Resting energy expenditure was measured by indirect calorimetry in 632 females and 148 males with severe obesity from the Utah Obesity Study. A literature search was conducted to identify prediction equations designed from, validated, or commonly used in samples with severe obesity. All equations were tested on each participant. Equations were considered unbiased if average predicted REE did not differ significantly ( p >0.05) from measured values. Bland-Altman plots characterized bias across measured REE values for prediction equations consistent with measured values. Precision was the percentage of the sample where an equation estimate was within 10% of measured REE. Equations were further assessed within sex and body mass index (BMI) subgroups. Results Only the body weight-based Lazzer equations (Lazzer A) and the Horie-Waitzberg equation generated unbiased prediction across all subgroups, with bias values ranging from -68.1 to 71.6 kcal, yet Bland-Altman plots revealed systematic bias, particularly at extreme values of REE. Equations including body composition did not predict better than body weight-based equations, and no single equation predicted best in every subgroup. Precision measurements never rose above 67.8%. Conclusions Clinicians may benefit from tailoring their choice of REE prediction equation to the specific characteristics of each patient, favoring equations with lower bias and greater precision within relevant subgroups. However, due to the low precision of REE prediction equations and the systematic bias revealed at REE extremes, it is highly recommended to measure REE whenever possible.

Source: Current Developments in Nutrition

DOI: 10.1016/j.cdnut.2025.107538