Longitudinal Trajectories of Resting Energy Expenditure, Cortisol and IGF-1, and Disease Severity in Critically Ill...

Background/Objectives: Resting energy expenditure (REE) rises during critical illness, but its determinants and prognostic relevance remain incompletely characterized. Identifying a circulating signal that tracks REE could provide a bedside surrogate for metabolic demand where indirect calorimetry is unavailable. This prospective pilot study described longitudinal trajectories of indirect calorimetry-measured REE, Sequential Organ Failure Assessment (SOFA) score, serum cortisol and insulin-like growth factor-1 (IGF-1) over two weeks in the intensive care unit (ICU), examined whether the REE trajectory was attenuated by adjustment for these variables, and generated preliminary effect-size and variance estimates for mortality. Methods: This single-center pilot study enrolled 39 critically ill adults at Evangelismos General Hospital (Athens, Greece); sample size was pragmatic, not based on an a priori power calculation. REE (Q-NRG metabolic monitor), cortisol and IGF-1 were measured at admission and days 5-7, 10-11 and 13-14, alongside SOFA. Trajectories were modeled with linear mixed-effects models using all available repeated measures. Confounding was assessed by adding SOFA, cortisol and IGF-1 as covariates and evaluating attenuation of the time effect; no formal mediation analysis was performed. Causes of missing data were quantified and a completers-only sensitivity analysis was undertaken. Mortality was analyzed by Cox regression with ICU length of stay as the time variable. Estimation was emphasized throughout; point estimates and 95% confidence intervals (CIs) are reported in preference to significance testing. Results: Mean age was 54.6 ± 18.1 years; 69.2% were male; median admission SOFA was 6 (IQR 3-9). ICU and 28-day mortality were 20.5% (8/39) and 10.3% (4/39). REE/kg rose from 25.3 ± 3.7 to a peak of 27.2 ± 4.2 kcal/kg/day by days 10-11 (likelihood-ratio χ² = 17.2, p = 0.0006). Attrition was driven predominantly by discharge alive (18 of 23 patients missing at days 13-14) rather than death (n = 2), and the trajectory was preserved in a completers-only sensitivity analysis (χ² = 9.13, p = 0.028). Cortisol declined (21.1 to 13.4 μg/dL) and IGF-1 rose (80.0 to 105.4 ng/mL); SOFA was essentially unchanged. REE did not correlate with SOFA, cortisol or IGF-1 at any time-point, and the time effect was not attenuated by adjustment for them. Admission REE was not associated with ICU mortality (HR 1.00, 95% CI 1.00-1.00). Higher admission REE was associated with a longer ICU stay, and this persisted after adjustment for body mass index (p = 0.026) and fat-free mass (p = 0.001). Age (HR 1.04, 95% CI 1.00-1.09) and admission SOFA (HR 1.18, 95% CI 0.98-1.44) were the covariates most associated with ICU mortality. Conclusions: REE rose progressively over the first 10-11 days of critical illness. Its trajectory was not attenuated by adjustment for SOFA, cortisol or IGF-1, which is compatible with-but does not establish-independence from the adrenal and somatotropic markers measured here. Neither REE nor these hormones were associated with mortality; age and disease severity remained the dominant prognostic factors, supporting a larger confirmatory study.