Enteral nutrient (EN) deprivation slows gastric emptying (GE) and increases plasma cholecystokinin (CCK) concentrations in healthy humans and may potentially contribute to the delayed GE in the critically ill. This study examined the impact of delayed feeding on GE, plasma CCK, and peptide YY (PYY) concentrations in the critically ill.Design:
Randomized controlled trial.Setting:
Mixed medical and surgical intensive care unit (ICU).Interventions:
Twenty-eight critically ill patients were randomized to receive EN either within 24 hrs of admission (“early feeding”: 54.9 ± 3.3 yrs; Acute Physiology and Chronic Health Evaluation (APACHE) II = 23.0 ± 1.8) or on day 4 of admission after GE assessment (“delayed feeding”: 56.1 ± 4.2 yrs, APACHE II = 21.7 ± 1.8). GE of 100 ml of Ensure was measured using scintigraphy on day 4 of admission. Blood was sampled for measurement of plasma CCK, PYY, and glucose concentrations.Results:
Demographics, APACHE II score, use of inotrope and morphine sedation were similar between the groups. The mean administered/prescribed caloric ratio in the “early feeding” group was 72 ± 4%. There were no differences in the retention of meal, intragastric meal distribution, proportion of patients with delayed GE (9/14 vs. 9/14), and plasma CCK and PYY concentrations during fasting and postprandially between the two groups. There was no relationship between the number of calories received and percentage of meal retention at 240 min (p > .05). However, delayed feeding was associated with longer duration of mechanical ventilations (13.7 ± 1.9 vs. 9.2 ± .9 days, p = .049) and length of stay in ICU (15.9 ± 1.9 vs. 11.3 ± 0.8 days, p = .048), but no difference in mortality.Conclusions:
In critical illness, delayed enteral feeding appears to have little impact on either GE or the enterogastric feedback hormones. However, the association between delayed feeding and increased duration of ventilation and length of stay in the ICU supports the current recommendation that enteral nutrition should be commenced early.