Intravenous amino acids may mediate the adverse effect of early parenteral nutrition on mortality in critically ill patients requiring mechanical ventilation: A post hoc analysis of the NEED trial

Abstract Background There is controversy regarding the optimal timing of initiating parenteral nutrition (PN) in critically ill patients. We aimed to evaluate the association between early PN and clinical outcomes and explore the mediation effects of different macronutrients in a cohort of mechanically ventilated patients. Methods This is a post hoc analysis of the NEED trial aiming to investigate the effect of implementing an evidence‐based feeding guideline in newly‐admitted critically ill patients. All eligible patients were divided into those who received early PN within the first 3 days of enrollment (early PN) or those who did not (non‐early PN). Propensity score matching with a one‐to‐one nearest neighbor‐matching algorithm was applied to control potential confounders. Mediation analysis was used to test the indirect effect of different macronutrients from PN on the relationship between early PN and 28‐day mortality. Results The propensity score matching created 370 matched pairs of 1154 patients that met the eligibility criteria. Compared with non‐early PN, patients receiving early PN had significantly higher 28‐day mortality (19.7% vs 12.4%; hazard ratio = 1.904; 95% CI, 1.063–3.410; P = 0.03). Mediation analysis showed that amino acids from early PN mediated 65% (mediation effect = 0.07; 95% CI, 0.02–0.13; P = 0.01) of the detrimental effect of early PN on the 28‐day mortality. Conclusion Early PN is associated with increased 28‐day mortality in critically ill patients requiring invasive mechanical ventilation. The detrimental effect may be mediated by intravenous amino acids from early PN.