Estimates of resting energy expenditure and total energy expenditure using predictive equations in adults with overweight and obesity: a systematic review with meta-analysis

• Published in: Nutrition reviews Vol. 80; no. 11; pp. 2113 - 2135
• Main Authors: Macena, Mateus de Lima, Paula, Déborah Tenório da Costa, da Silva Júnior, André Eduardo, Praxedes, Dafiny Rodrigues Silva, Pureza, Isabele Rejane de Oliveira Maranhão, de Melo, Ingrid Sofia Vieira, Bueno, Nassib Bezerra
• Format: Journal Article
• Language: English
• Published: United States 10-10-2022
• Subjects: calorimetry; predictive equations; doubly labeled water; energy metabolism; obesity
• ISSN: 0029-6643; 1753-4887
• DOI: 10.1093/nutrit/nuac031

Energy expenditure predictive equations can generate inaccurate estimates for overweight or obese individuals.
The objective of this review was to determine which predictive equations for resting energy expenditure (REE) and total energy expenditure (TEE) have the lowest bias and the highest precision in adults with overweight and obesity.
Searches were performed in January 2022 in MEDLINE, Web of Science, Scopus, CENTRAL, and the gray literature databases.
Meta-analyses were performed with equations included in more than 1 study. The DerSimonian and Laird random-effects model and the I2 statistic were used to quantify heterogeneity in the quantitative analyses. The Egger test was performed to assess potential publication biases, and metaregressions were conducted to explore the heterogeneity. Findings were presented separated by participants' body mass index classification (overweight and obesity).
Sixty-one studies were included. The FAO/WHO/UNU (1985) equation, which uses only body weight in its formula, showed the lowest bias in estimating REE (mean difference [MD] = 8.97 kcal; 95% CI = -26.99; 44.94). In the subgroup analysis for individuals with obesity, the Lazzer (2007) equation showed the lowest bias (MD = 4.70 kcal; 95% CI = -95.45; 104.86). The Harris-Benedict equation (1919) showed the highest precision values for individuals with overweight (60.65%) and for individuals with obesity (62.54%). Equations with body composition data showed the highest biases. The equation proposed by the Institute of Medicine (2005) showed the lowest bias (MD = -2.52 kcal; 95% CI = -125.94; 120.90) in estimating the TEE. Most analyses showed high heterogeneity (I2 > 90%). There was no evidence of publication bias.
For individuals with overweight, the FAO/WHO/UNU (1985) and the Harris-Benedict equations (1919) showed the lowest bias and the highest precision in predicting the REE, respectively. For individuals with obesity, the Harris-Benedict equation (1919) showed the highest precision and the Lazzer equation (2007) showed the lowest bias. More studies are needed on predictive equations to estimate the TEE.
PROSPERO registration no. CRD42021262969.