Year in review 2009: Critical Care - metabolism

Novel insights into the metabolic alterations of critical illness were published in Critical Care in 2009. The association between early hypoglycaemia/high glycemic variability and poor outcome was confirmed. Improvements in the understanding of the pathophysiological mechanisms of stress hyperglycemia and potential progress in the bedside management of glucose control were presented. With regard to enteral nutrition, some alterations of gastrointestinal physiology were better delineated. The relationship between the achievement of nutritional goals and outcomes was further investigated. Finally, understanding of some critical-illness-related endocrine and neuromuscular disorders improved through new experimental and clinical findings.


Glucose metabolism and control
Since 2001 and the publication of the 'Leuven I' study [1], the metabolism and control of blood glucose is an area of intense research in intensive care medicine. Th e contrasting fi ndings of subsequent prospective randomized controlled trials of glucose control fostered research in several diff erent fi elds: epidemiological insights came from association studies between blood glucose and outcome; endocrinological pathways were investigated as potential contributors to stress hyperglycaemia; and computer-assisted decision systems and continuous glucose monitoring were assessed in clinical conditions.

Association between blood glucose levels and outcome
An association between high admission blood glucose concentration and worse vital outcome has been known of for several decades. Interestingly, an analysis of a database of more than 250,000 patients allowed further categorisation of the type of patients in which the relationship between blood glucose and mortality was the strongest after acute myocardial infarction, arrhythmia, unstable angina or pulmonary embolism [2]. Likewise, the presence of hypoglycaemia at the time of admission or during critical illness is also associated with increased mortality, as reviewed recently [3]. An important article published in 2009 in Critical Care summarises the analysis of a 5-year database from the Australia New Zealand Intensive Care Society including 66,184 adult admissions [4]. Th ese authors indeed reported that the presence of early hypoglycaemia is common and associated with lower survival rates, especially in some subsets of patients. High early blood glucose variability (that is, during the fi rst 24 hours of ICU stay) was also associated with higher adjusted ICU and hospital mortality.

Physiopathology of stress hyperglycaemia
New important insights into the physiopathology of stress hyperglycaemia were also reported in Critical Care last year. In particular, the underlying mechanisms of insulin resistance were investigated by Langouche and colleagues [5] in a set of 318 critically ill patients. Th ese authors recorded the circulating blood levels of adiponectin, retinol binding protein 4 and leptin, three adipokines known to be involved in the modulation of insulin sensitivity and action. Th e levels of all three adipokines were lower than in normal subjects and intensive insulin therapy infl uenced the concentrations of adiponectin and retinol binding protein 4. Similarly, Venkatesh and colleagues reported in Critical Care [6] decreases in the mean plasma adiponectin concentration of 23 critically ill patients compared to historical control subjects. Interestingly, a correlation was found between adiponectin and plasma cortisol levels, and an inverse correlation was found between adiponectin and Creactive protein levels. Circulating levels of resistin, another recently reported hormone released by adipocytes and macrophages, were found to be increased in a population of 170 critically ill patients [7]; a correlation Abstract Novel insights into the metabolic alterations of critical illness were published in Critical Care in 2009. The association between early hypoglycaemia/ high glycemic variability and poor outcome was confi rmed. Improvements in the understanding of the pathophysiological mechanisms of stress hyperglycemia and potential progress in the bedside management of glucose control were presented. With regard to enteral nutrition, some alterations of gastrointestinal physiology were better delineated. The relationship between the achievement of nutritional goals and outcomes was further investigated. Finally, understanding of some critical-illness-related endocrine and neuromuscular disorders improved through new experimental and clinical fi ndings.
with the magnitude of insulin resistance was also reported. Th ese important fi ndings are quite consistent with recent progress in understanding the changes in adipose tissue gene expression during critical illness [8,9].
Another potentially important pathophysiological insight was reported by Preissig and colleagues in Critical Care in 2009 [10]. Th e pancreatic endocrine function of critically ill children admitted to a paediatric ICU with respiratory or cardiovascular failure was studied by the simultaneous analysis of blood glucose and plasmatic C-peptide levels. A severe primary beta-cell dysfunction refl ected by inappropriately low levels of C-peptide was found in children with both respiratory and cardio vascular failure; in contrast, high insulin resistance appeared as the prominent cause of stress hyperglycaemia in cases of respiratory failure only. Th ese intriguing fi ndings of diff erent and apparently unrelated patho genetic mecha nisms of stress hyperglycaemia prompted several diff erent interpretations of the adequacy of the beta-cell response [11,12].
In any case, impairments of glucose metabolism could diff er between children and adults. Th e important paediatric Leuven study published in 2009 by Vlasselaers and colleagues [13] confi rmed some benefi t of insulin therapy dosed to reach 'age-adjusted normoglycaemia' .

Clinical management of glucose control
Th e failure to reproduce the results of the Leuven I study [1] in diff erent settings, including two large-scale prospective randomised international trials published in 2009 [14,15], raised a number of important issues directly relevant to daily practice [16].
Th e two most recent meta-analyses [17,18] were unable to demonstrate an advantage of tight glucose control by intensive insulin therapy compared to a more conservative approach. As a consequence, the guidelines for glucose control were re-assessed in 2009 [19,20] and now recommend that blood glucose be kept below 180 mg/dl (10.0 mmol/L).
Whatever therapeutic target is chosen, quality of performance is a key factor needed to interpret the outcome variables of studies on tight glucose control [21]. Several approaches are currently under investiga tion, not only to improve the quality of glucose control, but also to increase safety and to decrease glycaemic variability. Th e use of continuous or near-continuous monitor ing devices and closed-loop systems for insulin infusion based on systematic algorithms based on multiple-compartment models are currently being tested. In Critical Care, Juneja and colleagues [22] report their experience of the use of a computerized insulin IV protocol program in 4,588 patients over 21 months. In essence, this study validated the effi cacy of this software as the target glycaemic range was rapidly achieved, with a very low rate of hypoglycaemia. However, frequent checks of glucose are required, implying an adapted nurse to patient ratio. Similar fi ndings were reported by others using other computer ized approaches [23][24][25]. Th e use of continuous blood glucose monitoring systems is another area of intense investigation [26,27].

Enteral nutrition and gastrointestinal disorders
Th e importance of early enteral nutrition was highlighted in 2009 by the publication of a meta-analysis in which a signifi cant reduction in mortality occurred when enteral nutrition was instituted within 24 hours after admission [28]. However, diffi culties in providing effi cient early enteral nutrition are frequent and prevent its institution in many situations. Clinical studies carried out and published in Critical Care in 2009 addressed the issues of digestive physiology during critical illness and the infl uence of enteral nutrition on outcome.

Digestive physiology
Chapman and colleagues [29] published a study concerning the relationship between gastric emptying, glucose absorption and glycaemia in critically ill patients. Using a 3-O-methyl-glucose test, they found that delayed gastric emptying decreased glucose absorption in a group of 19 mechanically ventilated patients compared to 19 healthy subjects; conversely, blood glucose concentration aff ected gastric emptying.
Th e relationship between digestive physiology and glucose was further investigated by Deane and colleagues [30]. Th e eff ects of exogenous glucagon-like peptide-1 on the glycaemic response to small intestinal nutrients was studied in critically ill patients. Glucagon-like peptide-1 is a hormone exerting an 'incretin' eff ect (that is, it stimulates insulin secretion), thereby off ering a novel therapeutic approach to reduce the magnitude of glycaemic responses during enteral feeding.

Enteral nutrition and clinical outcome
Of the unresolved issues regarding enteral nutrition, the best site of infusion has still to be determined. White and colleagues [31] addressed this question by comparing the outcomes of ventilated critically ill patients randomised to early post-pyloric or gastric feeding. In essence, early post-pyloric feeding did not provide any advantage over gastric feeding but did delay the achievement of the nutritional target. In terms of outcomes, there were no diff erences between the two groups.
A relationship between the achievement of nutritional goals and vital outcome was investigated by Strack van Schijndel and colleagues [32] in a prospective obser vational cohort study of 243 ventilated critically ill patients. Th e caloric and protein targets were analysed separately: enteral nutrition was calculated according to indirect calori metry and the protein intake was 1.2 g of protein/kg/ day. Reaching the nutritional goals was associated with signifi cant decreases in ICU and 28 day mortality in the female population, while the diff erence was not significant in males. Th ese challenging fi ndings are of interest when designing interventional trials. Similarly, another observational study reported an association between energy and protein intake and vital outcome in subsets of obese and lean patients [33].
Th e addition of pharmaconutrients is another area of research. For instance, glutamine could be of particular interest in cases of sepsis as it could attenuate infl ammation and increase the heat-shock protein response. Th ese pathways were investigated in healthy volunteers receiving endotoxin [34]. Even though endotoxin reduced plasma glutamine concentrations, there was no measurable eff ect on cytokine levels, nor on the expression of heatshock proteins.

Thyroidal and neuromuscular alterations
Both endocrine and neuromuscular changes are now identifi ed as important contributors to the poor functional outcome of patients with prolonged critical illness. For instance, the 'low T3 syndrome' has been recognised for a long time and its involvement in the need for prolonged mechanical ventilation has recently been revisited [35], although the pathophysiology of this frequent disorder is still incompletely elucidated. Mebis and colleagues [36] investigated the expression of genes in the hypothalamus of an animal model of chronic critical illness and found decreased mRNA levels of thyrotropin releasing hormone and increased mRNA levels of type II diodinase and thyroid hormone trans porters.
Th e neuromuscular consequences encountered in patients with prolonged critical illness are responsible for several complications [37], including dysfunction of the diaphragm muscle [38]. Th e pathophysiology of ICUacquired weakness or critical illness polyneuromyopathy is complex and involves increased oxidative stress, impaired microcirculation, proteolytic status, cytokinerelated infl ammation, altered calcium homeostasis, and hyperglycaemia [39]. Hermans and colleagues [40] published a study in Critical Care that focused on electrophysiological data from patients in their ICU before and after implementation of intensive insuline therapy. In their experience, intensive insulin therapy reduced the electro physiological incidence of critical illness polyneuro myopathy and the duration of mechanical ventilation. Th is eff ect was not found for other thera peutic modalities [41]. Th e potential mechanisms are partially speculative and were discussed in Critical Care [42].

Conclusion
Th e area of critical-illness-associated metabolic and endocrine changes received increased attention in 2009, as refl ected by the articles published. Th e issues of stress hyperglycaemia and glucose control were also further investigated. Th e nutritional aspects of critical illness, particularly derangements of the gastrointestinal physiology and the neuromuscular changes found in critically ill patients, were re-addressed using new approaches. Altogether, new areas of research were opened by the high-quality articles published in Critical Care in 2009.