The French intensive care societies (SRLF and SFAR), in agreement with the European and American societies for enteral and parenteral nutrition, recommend to quickly administer an artificial nutrition to patients admitted to ICU and for which it is expected that they will not be able to eat normally in the three days of admission.

Enteral nutrition should be used in priority if the gut is functioning. However, intolerance to enteral nutrition, such as vomiting, regurgitation, increased residual gastric volume, or diarrhea, occurs in 40% of patients hospitalized in ICU receiving enteral nutrition. Intolerance to enteral nutrition leads to the risk of not receiving enough nutrition. Feeding intolerance also exposes to the risk of acute mesenteric ischemia, especially in the most severe patients under catecholamine for shock. Currently, it is not possible to predict intolerance to enteral nutrition in ICU patients. Thus, the diagnosis of intolerance is made a posteriori while enteral nutrition is in progress.

Citrullinemia (normal concentration of 20 to 60 μmol / L), could be a good biomarker of the function of enterocytes involved in the absorption of food.

The aim of this study is to evaluate the interest of citrullinemia to predict tolerance to enteral nutrition in ICU patients.

Citrulline is an amino acid mainly produced by enterocytes from glutamine.

Citrullinemia ≤ 20μmol/L is a biomarker of enterocyte mass reduction, secondary to the reduction of small bowel length or secondary to villous atrophy diseases. A patient with low citrullinemia is probably at risk of enteral nutrition malabsorption.

Citrulline generation test consists of measuring the course of citrullinemia after an oral or intravenous administration of glutamine, the main precursor of citrulline. When the small bowel is functional, administration of glutamine is followed by the elevation of citrullinemia. On the contrary, when there is enterocyte dysfunction or reduction, citrullinemia does not vary after glutamine administration. Peeters et al have shown the feasibility of carrying out the citrulline generation test in patients hospitalized in the ICU. In this context, intravenous administration of glutamine (20 gr of 10% N2-L-Alanyl-L-Glutamine, dilution in 200 ml of water for injection, IV infusion over 30 minutes) followed by monitoring of citrullinemia (basal time and then after 90 minutes, on arterial sampling) is feasible and reproducible.

There is currently no data available on the relationship between citrullinemia measured at ICU admission and tolerance to enteral nutrition in the days following ICU admission. Similarly, the value of the citrulline generation test for evaluating enteral nutrition tolerance is unknown.

The investigators hypothesize that patients with low citrullinemia at ICU admission, as well as those with a decreased citrullinemia elevation after Dipeptiven bolus, will have less tolerance to enteral nutrition than patients with normal citrullinemia.

PREDICT is the first study evaluating the interest of citrullinemia and the citrulline generation test to predict feeding tolerance among critically ill patients.

Inclusion Criteria:

• 18 years old or older
• patient admited to ICU
• expected hospital stay of at least 3 days
• indication of enteral nutrition.
• presence of an arterial catheter
• signed informed consent

Exclusion Criteria:

• contraindication to enteral nutrition (e.g. non-functional digestive tract, mesenteric ischemia, occlusion, digestive fistula, active gastrointestinal bleeding)
• pregnancy ;
• chronic renal failure;
• chronic intestinal pathology