The optimal nutritional support regimen for critically ill patients remains a conundrum. Critically ill patients are at particularly high risk for the development of malnutrition, which is associated with a poor clinical outcome. The use of specialized nutritional support is widely believed to improve clinical outcome in these patients. Unfortunately, there are little objective clinical data to support this belief. Enthusiasm for the putative clinical benefits of nutritional therapy must be tempered by the recognition that there are inherent risks associated with its use. The provision of specialized nutritional support in critically ill patients needs to be carefully evaluated in order to maximize its benefits while minimizing its risks.

Ideally, nutritional regimens for critically ill patients would be based on the results of high-quality clinical investigations in appropriate patient populations. Unfortunately, such evidence is limited, so there are wide variations in the practice of providing specialized nutritional support to critically ill patients. Within the context of the available evidence, it appears that the risks associated with the administration of nutrients beyond metabolic needs can easily outweigh the potential benefits of providing specialized nutritional support. In contrast, there is little evidence that use of hypocaloric nutritional regimens, at least for the short term, results in worse outcomes than use of standard nutritional regimens. Thus, the optimal nutritional regimen for critically ill patients may be one that minimizes the chance of overfeeding, even if it results in underfeeding a significant number of patients.

What is the evidence supporting less aggressive nutritional regimens? Outside of the hospital setting, the predominant nutritional problem faced by residents of developed countries is excessive energy and nutrient intake, resulting in obesity. In healthy outpatients, short periods of hypocaloric intake are well tolerated and result in improved insulin sensitivity and plasma lipid profiles. Although critically ill patients differ physiologically from unstressed patients, maintaining the excess weight of overweight or obese ICU patients receiving specialized nutritional support would not seem to be inherently beneficial. A common misconception is that acutely ill patients will lose large amounts of lean body mass if they do not receive a caloric load equal to or greater than their metabolic demands, but there is little evidence to support this belief. Compared to full caloric supplementation, mild to moderate caloric restriction does not result in further nitrogen losses among patients receiving specialized nutritional support, particularly if adequate amounts of nitrogen are administered.

In contrast, it is increasingly recognized that even brief periods of overfeeding have potentially adverse consequences; this may be particularly significant for critically ill patients. Short periods of hyperglycemia, which may result from excess carbohydrate supply, are associated with a significant increase in infectious complications, a major cause of mortality in critically ill patients. Excess caloric supply also increases triglyceride production and may result in significant hypertriglyceridemia, causing pancreatitis in extreme cases. In addition, increased CO2 production caused by overfeeding may lead to prolonged mechanical ventilation and its attendant consequences.

A common practice is to use a single formula, for instance a caloric intake of 25 kcal/kg, for all patients. However, measured energy expenditures vary widely in critically ill patients, even among those suffering from similar pathologic disorders.1 The measured energy expenditure of a patient is often lower than the estimated basal energy expenditure. Among ventilated patients at a long-term acute care facility, for instance, 20% had such low energy expenditures, and two-thirds were overfed with a “standard” nutritional prescription of 25 kcal/kg.2

Thus, the use of standard nutritional regimens may result in overfeeding significant numbers of patients, and so-called hypocaloric regimens might actually meet the nutritional needs of many critically ill patients. But would routine use of hypocaloric regimens cause harm to patients with higher caloric demands? Reassuringly, most of the clinical evidence indicates that this approach would not adversely affect clinical outcome.

Indirect evidence can be inferred from previous studies of specialized nutritional support, particularly meta-analyses of such studies. For instance, routine use of parenteral nutrition (PN) in perioperative patients was generally not beneficial unless it was used in patients with severe, underlying malnutrition.3 In trials comparing enteral nutrition (EN) to PN, patients randomized to receive PN received more macronutrients than patients randomized to EN but did not have any clinical benefits attributable to greater nutrient delivery.4 Among patients receiving EN only, a retrospective study also found no difference in outcome among patients who received fewer calories.5

Most prospective clinical trials comparing conventional PN with hypocaloric PN find no disadvantage to the hypocaloric regimen. In two small trials of obese surgical patients receiving PN and another small trial of trauma patients, no significant decreases in nitrogen retention were noted with the use of the hypocaloric regimens.6-8 Unfortunately, these trials did not specifically address clinical endpoints. In a trial comparing lipid-free, hypocaloric PN to standard PN in trauma patients, there were significantly fewer infections, ventilator days, ICU days and hospital days in patients receiving the hypocaloric regimen.9 Another study comparing standard to hypocaloric PN found no clinical benefit to the hypocaloric regimen, although infectious complications were insignificantly reduced in this group.10 In this study, nitrogen losses were greater in patients on the hypocaloric regimen, but these patients received lower amounts of amino acids. Interestingly, patients randomized to the standard PN regimen actually received, on the average, only 1,200 kcal/day (18 kcal/kg/day) during the study period, which for many practitioners would represent a hypocaloric regimen.

Overall, there are significant uncertainties regarding the optimum caloric regimen for critically ill patients. It is clear that there are very real complications associated with overfeeding the critically ill patient, and there is reasonable evidence that underfeeding is not harmful. A rational approach at present would be to provide no greater than 20 total kcal/kg for patients with a normal body mass index and correspondingly lower amounts for overweight and obese patients. Higher caloric support should be restricted to critically ill patients with severe preexisting malnutrition. Well-designed trials are obviously needed to determine if this approach improves clinical outcome in critically ill patients.