The GH Axis in Relation to Accepting an Early Macronutrient Deficit and Outcome of Critically Ill Patients.

CONTEXT

Changes in the GH axis during critical illness resemble fasting in healthy adults and contribute to hypercatabolism, which potentially affects outcome. Accepting macronutrient deficits by withholding parenteral nutrition (PN) during the first week in the intensive care unit (ICU; late PN) reduced complications and accelerated recovery as compared with early use of PN (early PN).

OBJECTIVE

To investigate how late PN affects the GH axis in relation to its clinical outcome benefits.

DESIGN

Preplanned subanalysis of the Early Parenteral Nutrition Completing Enteral Nutrition in Adult Critically Ill Patients randomized controlled trial.

PARTICIPANTS

A total of 1128 patients for time-course study, 20 patients investigated for nocturnal GH pulsatility, and 600 patients investigated for muscle weakness, with early PN and late PN patients having comparable baseline characteristics.

INTERVENTION

Withholding PN during the first ICU week (late PN) vs early PN.

MAIN OUTCOME MEASURES

Changes in serum GH, IGF-I, IGF-binding protein (IGFBP) 3, and IGFBP1 concentrations from ICU admission to day 4 or last ICU day for patients with a shorter ICU stay (d4/LD) and association in multivariable analyses with likelihood of earlier live ICU discharge, risk of new infection, and muscle weakness.

RESULTS

Late PN attenuated a rise in serum GH and IGF-I (P < 0.0001), did not affect IGFBP3, and attenuated a decrease in IGFBP1 concentrations from admission to d4/LD (P < 0.0001) as compared with early PN. Late PN decreased nonpulsatile (P = 0.005), but not pulsatile, GH secretion. Adjusting the multivariable models for the observed GH axis alterations increased the independent benefit of late PN for all outcomes. GH axis alterations induced by late PN were independently associated with adverse outcomes (P ≤ 0.03).

CONCLUSION

Accepting macronutrient deficits early during critical illness further suppressed the GH axis, which statistically attenuated its clinical outcome benefits.