Comparing indirect calorimetry and predictive equations for energy expenditure - 1 year follow-up of adult hematopoietic stem cell transplanted recipients.

Abstract

Background and aims: Recipients of allogeneic hematopoietic stem cell transplantation (allo-HSCT) often undergo body composition changes that may impact their energy expenditure (EE), however, the accurate energy requirements may be difficult to estimate. Indirect calorimetry (IC) is recommended as the gold standard to assess measured energy expenditure (mEE), but its use is resource-demanding and thus rarely applied compared to predictive energy equations. Here we evaluated the agreement between mEE by IC and EE estimated by several predictive energy equations, or the ESPEN recommendation, during 1-year follow-up of adult allo-HSCT recipients.

Methods: We examined 117 adult patients with hematological malignancies who received allo-HSCT. mEE was assessed by IC in non-fasting patients before transplantation, and at 3, 6, 9 and 12 months post-allo-HSCT. Agreement between mEE and predictive values using the Harris-Benedict-HB, Mifflin St. Jeor-MSJ, Ireton Jones-IJ equations and the fixed ESPEN recommendation of 25 kcal/kg/day was evaluated using Bland-Altman analysis and accuracy within ± 10%. Changes in mEE over time were analyzed using linear mixed model with age as a covariate.

Results: A total of 509 mEE measurements by IC were obtained. HB and MSJ overestimated mEE on average (mean bias: + 2.9 and + 3.8 kcal/kg/day), while IJ, and the fixed ESPEN recommendation underestimated mEE (biases: - 0.7 and - 4.7 kcal/kg/day, respectively). None of the equations or the fixed ESPEN recommendation achieved > 50% accuracy within ± 10% of mEE by IC. IJ had the highest variability and widest limits of agreement. mEE changed significantly over time (p < 0.001), with no significant differences between the initially randomized intervention and control group (p = 0.982). Mean mREE was 24.74 ± 3.80 kcal/kg/day and 31.14 ± 4.91 kcal/kg/day when adjusted for FFM. Age was not significantly associated with mREE (p = 0.198).

Conclusion: Our findings indicate that commonly used predictive equations and the fixed ESPEN recommendation showed limited accuracy compared with mEE in allo-HSCT recipients. Only mEE obtained by IC captured longitudinal changes in EE during the first year post-transplantation. mEE by IC varied significantly the first year post-transplant.

Clinicaltrials: gov, ID NCT01181076.