None of the commonly used predictive equations provides high individual accuracy for estimating REE in adults with obesity, with mean absolute percentage error exceeding 15% in more than 50% of patients, and accuracy declining with increasing age and obesity severity.
Key Findings
Results
All analyzed predictive equations demonstrated substantial individual error in estimating resting energy expenditure in adults with obesity.
Mean absolute percentage error exceeded 15% in more than 50% of patients for all equations tested.
The study included 293 patients (111 men and 182 women) with a median BMI of 38.3 [32.9; 44.0] kg/m².
Median REE measured by indirect calorimetry was 1964.5 [1570.8; 2370.5] kcal/day.
Accuracy was evaluated using MAE, MAPE, RMSE, Pearson correlation coefficient, and Bland-Altman analysis.
Results
The Roza-Shizgal, WHO (Schofield), and Harris-Benedict equations showed the smallest mean bias relative to indirect calorimetry-measured values.
Despite having the smallest mean bias among the eight equations tested, these formulas still showed wide limits of agreement in Bland-Altman analysis.
Eight commonly used predictive equations were compared against indirect calorimetry (Cosmed K5) as the gold standard.
Wide limits of agreement indicate pronounced interindividual variability even for the best-performing equations.
Results
The accuracy of all predictive equations declined with increasing age and obesity severity.
Accuracy was lowest in patients with BMI ≥40 and BMI ≥50 kg/m².
Even equations with minimal mean bias were characterized by pronounced interindividual variability and reduced accuracy with increasing age and obesity severity.
The study population had a median BMI of 38.3 [32.9; 44.0] kg/m², providing a range of obesity severities for comparison.
Results
Equations based on fat-free mass systematically underestimated REE in patients with morbid obesity.
This systematic underestimation was specific to patients with morbid obesity.
Body composition was assessed by bioelectrical impedance analysis using InBody770.
Fat-free mass-based equations performed differently from anthropometric equations in this subgroup.
Conclusions
Indirect calorimetry remains the most reliable method for determining energy requirements, particularly in patients with severe and morbid obesity.
The study was an open cross-sectional comparative design conducted in adults with BMI ≥25 kg/m².
Indirect calorimetry using Cosmed K5 was used as the gold standard comparator.
No predictive equation achieved high individual accuracy, supporting IC use in clinical practice for this population.
Although IC is considered the gold standard, predictive equations are predominantly used in clinical practice despite their limited and variable accuracy in individuals with obesity.
What This Means
This research suggests that the mathematical formulas commonly used in clinical settings to estimate how many calories a person burns at rest are significantly inaccurate for people with obesity. The study measured actual resting energy expenditure in 293 adults with overweight or obesity using a specialized breathing device (indirect calorimetry), then compared those measurements to predictions from eight widely-used formulas. More than half of patients had prediction errors greater than 15% for every formula tested, meaning a patient truly burning 2000 calories per day could routinely be estimated at anywhere from 1700 to 2300 calories or more.
The formulas performed even worse as patients got older or heavier, with the greatest inaccuracies seen in patients with a BMI of 40 or above. Formulas that use muscle mass measurements tended to underestimate calorie burn in people with the most severe obesity. While three formulas — Roza-Shizgal, WHO/Schofield, and Harris-Benedict — came closest on average, they still showed wide individual variation, meaning they might work reasonably well for a group but could be quite wrong for any given person.
This research suggests that relying on predictive equations to set calorie targets for people with obesity carries meaningful risk of error, which could undermine weight management efforts. The findings highlight the practical value of measuring resting energy expenditure directly through indirect calorimetry, especially for patients with severe or morbid obesity, where the stakes of miscalculation are highest and the inaccuracies of standard formulas are most pronounced.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Averkina A, Rafaelyan M, Vasyukova O, Guseinova R, Okorokov P, Burmitskaya Y, et al.. (2026). [Resting Energy Expenditure Assessment in Adults with Obesity: Limitations of Commonly Used Predictive Equations].. Problemy endokrinologii. https://doi.org/10.14341/probl13703