aDivision of Endocrinology, Diabetes and Bone Disease, Mount Sinai School of Medicine, New York, USA
bDepartment of Intensive Care Medicine and Burns Centre, University Hospital (Centre Hospitalier Universitaire Vaudos, CHUV), Lausanne, Switzerland
Correspondence to Jeffrey I. Mechanick, MD, 1192 Park Avenue, New York, NY 10128, USA E-mail: firstname.lastname@example.org
For the last decade, we have discussed several controversies in the field of metabolic support in the ICU, many of which characterized by vacillations regarding tight glycemic control or relaxed targets, thresholds for underfeeding and overfeeding, and the use of enteral versus parenteral nutrition. The impetus, or driving force, propelling these pendulums of discord has been both theoretical and empirical. However, this year has seen meta-analyses, clinical trials, and derivative position papers that have diminished the swing of these pendulums. For instance, although Hermanides et al.  have corroborated previous findings  that increased glycemic variability can be a negative prognosticator in the ICU, Meyfroidt et al.  have demonstrated that this parameter did not primarily account for the benefit of intensive insulin therapy (IIT) in the two landmark Leuven studies [4,5]. However, as pointed out very well by Dr Bistrian in an editorial comment , the routine concomitant use of early nutrition support, mainly via the parenteral route in the first SICU Leuven study , may have produced less glycemic variability then in the second MICU Leuven Study  in which there may have been more frequent enteral nutrition interruptions after the first 3 ICU days. Consequently, Meyfroidt and Van den Berghe  re-examined their data and these suspicions were confirmed in addition to the observation that enteral nutrition, perhaps also due to increased glycemic variability with multiple interruptions, was associated with an increased risk for hypoglycemia; the latter may also be partly due to insufficient enteral feeding in several trials . In fact, this concept of marrying more aggressive – early combined parenteral and enteral – nutrition support with IIT targeting a BG of 80–110 mg/dl to achieve better clinical ICU outcomes, with less hypoglycemia than in patients who did not receive parenteral nutrition, was demonstrated by a meta-analysis by Marik and Preiser . Clear supporting arguments by Blackburn et al.  and Bistrian  followed. The net paradigm of metabolic support necessarily incorporating both parenteral nutrition and IIT was articulated by Burke et al.  and Scurlock et al. [13,14]. Hence, the stage is set once again for the present issue in which the following articles are presented within the context of this new combined-modality metabolic support paradigm.
Confronted with the continued worldwide hospital and ICU underfeeding epidemics  and with conflicting papers about the eventual benefits of hypocaloric feeding, it was important to provide a physiologic background for the energy requirements during critical illness. In this issue, Fontaine and Müller (pp. 171–175) set the scene of cellular energetic. In a pleasant easily readable text, the authors show that feeding is like breathing, both being vital, only with a different time constant. This explains why critical care teams may be less sensitive to underfeeding: the deleterious effects of underfeeding take a relatively long time to become manifest, whereas the effects of severe hypoxia become obvious within minutes. Similar to underfeeding, the deleterious effects of overfeeding ICU patients should also be avoided. Thus, an important problem arises: what is a ‘scientifically sound energy target’. Total energy expenditure is highly variable in the ICU patient. Indirect calorimetry remains the only practical method to quantitatively guide feeding but it is not available in the majority of ICUs. As a result, the current clinical practice guidelines on artificial nutrition should be applied: the recommended targets (20–25 kcal/kg per day) have been shown in numerous well conducted metabolic studies to be sound. This rule of thumb is also globally applicable in the elderly patient, whose basal metabolism is reduced: it would only rarely cause significant overfeeding. The parallel with hypoxia is important from another point of view: feeding should be started early: volunteer studies have shown that a 12-h fast causes alterations not only of glycogen liver content, but also of mitochondrial function .
The critically ill burn patient is ‘extreme’ in many ways. In particular, the inflammatory, metabolic and endocrine responses are grotesque compared to the vast majority of ICU patients. In this issue, Gauglitz et al. (pp. 176–181) provide an up-to-date summary of metabolic and pharmacologic tools available for use with burn patients. Results that are derived from burn patients apply to other conditions and pharmacologic modulators of the hypercatabolic response to critical illness have emerged. The attenuation of the hypermetabolic response by various pharmacologic modalities has become an essential component of the management of severe burn patients. Other pharmacological strategies include the use of growth hormone (GH), insulin-like growth factor, oxandrolone, and IIT. Although the administration of GH cannot be advocated in adult critically ill patients after the increased death rate observed in two large European trials , its use in burned children who suffer major downregulation of their hypothalamic–pituitary axis may be warranted. The authors also discuss promising alternative strategies, such as using glucagon-like-peptide (GLP)-1 and the PPAR-γ agonists. The ‘burns laboratory’ is likely to provide many answers in the future.
In the article by Peterson et al. (pp. 182–185) on oral nutrition in the ICU, there are two points worth focusing on. The first point resonates with data from other observational studies and clinical trials in the ICU that simple underfeeding, common when oral nutrition is relied on exclusively, is associated with poorer outcomes. However, the second point is that among those ICU patients who are not being mechanically ventilated, there is very little evidence to provide guidelines for management with oral nutrition. Our suspicion is that, to a large extent, this is a ‘cultural’ problem in the ICU wherein medical and nursing personnel prefer not to be aggressive in placing enteral access devices or starting parenteral nutrition in the hope that oral nutrition will spontaneously recover. Current clinical practice guidelines reflect this approach by the continued inclusion of waiting periods before nutrition support is started. Although this may be appropriate for lower risk patients, controlled trials are needed that stratify risk and specifically examine the role of oral nutrition methods.
In the article by Byrnes and Stangenes (pp. 186–192), the refeeding syndrome is reviewed as it applied to patients in the ICU. Although the electrolyte abnormalities (mainly hypophosphatemia), micronutrient redistributions, and cardiac problems are similar to noncritically ill patients, it is the respiratory failure from diaphragmatic fatigue that poses the greatest threat in the ICU. Nutrition support in patients at high risk for refeeding syndrome needs to be quantitatively less initially, in terms of nonprotein calories, with slow progression to target amounts. However, questions still remain: how to best predict patients who will develop refeeding syndrome in the ICU, what is the optimal formula for protein–calorie–micronutrient refeeding, and what are the specific glycemic targets with IIT? Let us reasonably guess that 4–5 days without nutrition in the hospital are probably enough to cause acute underfeeding and thereby increase the risk of developing the refeeding syndrome.
In the article by DeLegge (pp. 193–196), the myth of needing to hold enteral nutrition with large gastric residual volumes (GRVs) is critically examined in terms of the clinical evidence. This is important because enteral nutrition interruptions are responsible for increased glycemic variability and protein–calorie underfeeding, both of which independently have been associated with poorer clinical outcomes. It is unclear whether monitoring GRVs will decrease the likelihood of aspiration pneumonias. It is also unclear whether the empiric use of prokinetic agents will reduce GRVs.
How should a critically ill patient with upper gastrointestinal bleeding, and the one with nonbleeding esophageal varices be fed? Are there any reasons for withholding enteral nutrition? Are there objective reasons for giving parenteral nutrition to all? In this issue, Hébuterne and Vanbiervliet (pp. 197–201) show us that the evidence, albeit scarce, favors providing enteral nutrition. The authors show the differences between patients who develop gastrointestinal bleeding, while in the ICU, and those admitted for acute gastrointestinal bleeding for either peptic ulcer or esophageal varices due to portal hypertension. The available data show that whatever the cause of the bleeding, enteral/oral feeding can (or should) be resumed within 48 h of endoscopic therapy and cessation of bleeding: further the use of fine bore feeding tubes has shown to be safe in the presence of esophageal varices. We should simply apply these guidelines.
In the article by Berger and Que (pp. 202–208), the role of computerized information systems (CISs) in coordinating the necessary use of both nutrition support and IIT in the management of critically ill patients is illustrated. The authors successfully argue that optimal delivery of these two interventions will require advanced bioinformatics and that with future economic pressures on metabolic support, CISs will likely become mandated.
In an aging population, with increasing economic pressures and an imperfect healthcare system, curtailing costs is urgently required. Nutrition therapy has invariably been considered by our administrators as cost-generating: administrators have trouble considering that this investment may bring positive economic results. Indeed, cost-efficiency studies are difficult to conduct, as the inclusion of all of the dependent and independent variables in the equations is tricky. In this issue, Scurlock et al. (pp. 209–212) analyze the economical implications of glycemic control. Tight glycemic control clearly constitutes an additional cost that should be considered an investment: increased workload for the nurses, increased use of disposables or arterial blood gas determination, etc. do cost money. The authors show that IIT and metabolic care offers a cost-effective method to reduce complications: the avoidance of hyperglycemia is associated with a significant reduction of infectious complications such as mediastinitis after cardiac surgery. Several studies have confirmed the original Leuven results demonstrating that IIT reduces morbidity and mortality through reduction of the utilization of hospital resources (i.e., blood products, ventilator days, renal replacement therapy, and length of stay). In studies addressing this cost issue, published savings per patient benefiting from tight control vary between €2500 and US$ 5500.
In conclusion, the gradual resolution of controversy directs us toward a state of equilibrium. Namely, that critically ill patients need to be understood in terms of their risk stratification, disease process, and timeline of the immune-neuroendocrine stress response. Then, they receive metabolic support fashioned to incorporate both early, combined enteral and parenteral nutrition and IIT with caloric, protein, and glycemic targets commensurate with their overall state (‘metabolic support’). This means that for many critically ill patients, routine application of parenteral nutrition and glycemic targets of 80–110 mg/dl will be associated with favorable outcomes. Advanced bioinformatics to optimize the transduction of complex information into actions, as well as continued monitoring of the economic impact of these interventions will need to be implemented. The pendulums of controversy have not stopped, but their swings may now be slow enough for us to see these dilemmas more clearly and focus on emerging answers.
1 Hermanides J, Vriesendorp TM, Bosman RJ, et al
. Glucose variability is associated with intensive care unit mortality. Crit Care Med 2010; 38:838–842.
2 Krinsley JS. Glycemic variability: a strong independent predictor of mortality in critically ill patients. Crit Care Med 2008; 36:3008–3013.
3 Meyfroidt G, Keenan DM, Wang X, et al
. Dynamic characteristics of blood glucose time series during the course of critical illness: effects of intensive insulin therapy and relative association with mortality. Crit Care Med 2010; 38:1021–1029.
4 Van den Berghe G, Wouters P, Weekers P, et al
. Intensive insulin therapy in critically ill patients. N Engl J Med 2001; 345:1359–1367.
5 Van den Berghe G, Wilmer A, Hermans G, et al
. Intensive insulin therapy in medical intensive care patients. N Engl J Med 2006; 354:449–461.
6 Bistrian BR. Parenteral feeding and intensive insulin therapy. Crit Care Med 2010; 38:1922.
7 Meyfroidt G, Van den Berghe G. Author's reply. Crit Care Med 2010; 38:1922–1923.
8 Berger MM, Mechanick JI. Continuing controversy in the intensive care unit: why tight glycemic control, nutrition support, and nutritional pharmacology are each necessary therapeutic considerations. Curr Opin Clin Nutr Metab Care 2010; 13:167–169.
9 Marik PE, Preiser J-C. Toward understanding tight glycemic control in the ICU. Chest 2010; 137:544–551.
10 Blackburn GL, Wollner S, Bistrian BR. Nutrition support in the intensive care unit. Arch Surg 2010; 145:533–538.
11 Bistrian BR. Comment on ‘Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient’. J Parent Enteral Nutr 2010; 34:348–349.
12 Burke PA, Young LS, Bistrian BR. Metabolic vs nutrition support: a hypothesis. J Parent Enteral Nutr 2010; 34:546–548.
13 Scurlock CS, Raikhelkar J, Mechanick J. Parenteral nutrition in the critically ill patient. N Engl J Med 2010; 362:81.
14 Scurlock CS, Raikhelkar J, Mechanick JI. Intensive metabolic support: evolution and revolution. Endocr Pract 2008; 14:1047–1054.
15 Alberda C, Gramlich L, Jones N, et al
. The relationship between nutritional intake and clinical outcomes in critically ill patients: results of an international multicenter observational study. Intensive Care Med 2009; 35:1728–1737.
16 Awad S, Stephenson MC, Placidi E, et al
. The effects of fasting and refeeding with a ‘metabolic preconditioning’ drink on substrate reserves and mononuclear cell mitochondrial function. Clin Nutr 2010; 29:538–544.
17 Takala J, Ruokonen E, Webster NR, et al
. Increased mortality associated with growth hormone treatment in critically ill adults. New Engl J Med 1999; 341:785–792.