Resting Energy Expenditure in COPD: development of a new predictive equation

Background: Resting energy expenditure (REE) is elevated in patients affected by chronic obstructive pulmonary disease (COPD). Although indirect calorimetry (IC) represents the gold standard for its measurement, its cost and the need of qualified technicians limit its use, compelling clinicians to use predictive equations not validated in COPD, like the Harris-Benedict (HB), or significantly overestimating the REE, like the Moore-Angelillo (MA). Aim: to develop a predictive equation to estimate REE in COPD. Methods: Stable COPD patients were consecutively enrolled in the pulmonary outpatient clinic of Campus Bio-Medico Teaching Hospital. Participants underwent a multidimensional assessment including medical history and examination, spirometry, 6 minute walking (6MW) test, bioimpedance analysis, CT-scan and IC. The association of REE with each variable was assessed using linear regression. Variables were selected using a backward stepwise regression analysis. The new equation was obtained using a cross-validated lasso linear regression. HB, MA and the new equation were compared to REE using a Bland-Altman plot. Results: 61 COPD patients, mean age 72 (SD:7) years, underwent IC with a measured REE of 1662 Kcal (SD:335), significantly higher than HB REE [1460 Kcal (SD:242), p: 0.002] and lower than MA REE [1772 Kcal (SD:266), p: 0.04]. REE was associated with sex, height, fat-free mass index, dyspnea, exacerbations and comorbidities. The equation based on fat-free mass index, pack/year, oxygen saturation, and 6MW distance (cv-adjR2 0.42, p:<0.01) estimated a mean REE of 1677 Kcal (SD:244), improving REE prediction. Conclusions: The new equation outperforms HB and MA in predicting REE in COPD.