New equation to estimate resting energy expenditure in non-critically ill patients

Prediction of resting energy expenditure in Italian older adults with severe obesity

Background

In the last decade a large number of studies proposed and/or validated equations to estimate the Resting Energy Expenditure (REE) in adults and/or older adults, however, no equation currently available showed good accuracy for older adults with severe obesity. Thus, this study aimed to develop and validate new predictive equations for REE, based on data from the indirect calorimetry, in Italian older adults with severe obesity.

Methods

A retrospective study was as conducted with 764 Caucasian older adults with severe obesity (age range: 60-74 years and BMI ≥ 35 kg/m/²). Four models were used to test the accuracy of anthropometry and body composition variables in multivariable prediction of REE. All models were derived by stepwise multiple regression analysis using a calibration group of 382 subjects [295 females and 87 males] and the equations were cross-validated in the remaining 382 subjects [295 females and 87 males] as validation group. The new prediction equations and the other published equations were tested using the Bland-Altman method. Prediction accuracy was defined as the percentage of subjects whose REE was predicted within ± 10% of measured REE.

Results

All the equations analyzed predicted higher energy requirements for males than females, and most of them underestimated the energy requirement values of our sample. The highest accuracy values were observed in the new equations, with 62% in the anthropometric model and 63% in the body composition model.

Conclusion

Although the accuracy of our equations was slightly higher in comparison with the other taken into consideration, they cannot be considered completely satisfactory for predicting REE in Italians older adults with severe obesity. When predicting equations cannot guarantee precise or acceptable values of REE, the use of indirect calorimetry (if available) should be always recommended, especially in clinical practice.

Resting energy expenditure in cancer patients: Agreement between predictive equations and indirect calorimetry

BACKGROUND/OBJECTIVES

Disorders of energy metabolism is a common phenomenon in cancer patients. Changes in resting energy expenditure (REE) combined with inadequate nutrition support appear to be causes of nutritional depletion in cancer patients. In clinical practice, REE is typically calculated using predictive equations. The aim of this study was to determine the agreement between REE estimated by predictive equations and REE measured by IC in Portuguese cancer patients. Differences in measured REE between patients with different types of digestive cancers were also assessed.

SUBJECTS/METHODS

REE was measured by indirect calorimetry (IC) in 61 patients with cancer diagnosis (gastric cancer, cholangiocarcinoma, pancreatic cancer, liver cancer and colorectal cancer). Measured REE values were compared with those estimated by equations of Harris-Benedict, Schofield, Ireton-Jones, Mifflin-St.Jeor and Barcellos I and II.

RESULTS

Mean Respiratory Quotient (RQ) was 0.77 ± 0.09, which indicates high lipids utilization as substrate. No statistically significant differences between REE or RQ from patients with different cancer types were observed. All equations underestimate REE: Harris-Benedict, mean difference -648 kcal (limits of agreement +627 to -1923 kcal); MifflinSt.Jeor, mean difference -694 kcal (limits of agreement +544 to -193 kcal); Schofield, mean difference -531 kcal (limits of agreement +662 to -1723 kcal); and Ireton-Jones, mean difference -556 kcal (limits of agreement +774 to -1887 kcal). Barcellos I and II showed lower mean difference when compared to measured REE, +59 and + 52 kcal, respectively, although presenting wide limits of agreement, +1542 to -1424 kcal and +1429 to -1326, respectively.

CONCLUSIONS

Although Barcellos Equations underestimate less and enable more accurate average REE prediction in cancer patients, still present wide limits of agreement and therefore clinically important differences in REE estimation may be found at individual level. Our results support the appropriateness of measuring REE by IC to better adequate the nutrition support to cancer patients. Further research is needed to improve the current knowledge base of energy expenditure in cancer patients, and to improve the accuracy of existing predictive equations.