Answers are grounded in this paper and cite their sources.
Reviews metabolic responses to surgical trauma and nutritional therapies, highlighting the roles of glycemic control, indirect calorimetry, and gut immunity.
This section on perioperative nutrition appears to be very different from the material presented a year ago on the same general subject. However, on closer inspection, certain associations are evident in at least three areas. Route and amount of caloric supply: the first issue dealt extensively with a comparison of oral, enteral and parenteral routes for caloric intake. This issue deals with the amount of calories needed to achieve caloric balance rather than the particular route of administration. The metabolic response to illness and injury: the first issue considered the metabolic response in terms of salt and water while this issue shifts the focus to carbohydrate metabolism and insulin action. Nutrition and immunology: the first issue reviewed immunoenhancing agents under the title of nutritional pharmacology. This issue examines the relation of the gut to immune function with an emphasis on nutritional factors. Gas exchange and indirect calorimetry Indirect calorimetry is the method by which the type and rate of substrate oxidation and heat production are calculated from measurement of gas exchange. The development and interpretation of gas exchange has represented a fundamental milestone for chemistry, biology and medicine. Yet calorimetry continues to receive only limited attention in relation to patient care. During the period of 1925-1950, most hospitals established facilities for the measurement of basal metabolism to diagnose and treat thyroid disease. Ironically, the same period was one where the attention of nutritional scientists shifted away from energy expenditure to enzymology, vitamins and metabolic pathways. When chemical measurements of thyroid function became widely available by 1950, hospitals dismantled their basal metabolic rate measuring facilities. Thus an era began when patients were treated on the basis of an assumed energy expenditure which was never measured and sometimes never thought about [1]. There has been a resurgence of interest in indirect calorimetry since the 1970s which represents a confluence of various factors. The new interest in calorimetry moved from attention to thyroid disease, to the influence of food and physical activity on energy expenditure. Bedside measuring equipment became available which offered the convenience of making immediate calculations and presenting the measurement of gas exchange in terms of calories per unit of body size. Such instruments, when properly calibrated, offer convenience and efficiency. However, the measurement of energy expenditure still has not become a central part of patient care. This may be because changes in resting energy expenditure are not usually disease-specific as in thyroid disease. More importantly, most doctors do not think of energy expenditure because they do not feel that knowing such information can ‘make a difference’ in managing a patient. Further reasons why calorimetry is not of more clinical use, may be divided into: (1) prediction vs measurement, (2) interpretation, (3) extrapolation and (4) relevance or outcome. Is prediction of energy expenditure from published normal values just as satisfactory as a single measurement? There is growing evidence that equations, such as the Harris-Benedict equations may provide a reasonable approximation for relatively normal patients. However, the sicker the patient, the more their energy expenditure cannot be predicted with assurance and should be measured. The potential errors of measurement have been reviewed extensively [2,3]. The paper by McClave and associates (pp. 61-67) has presented further useful data on duration and timing of individual measurements during each day. There are many issues related to interpretation of gas exchange data including the following: (1) doubt about appropriate standards, since body weight may involve an abnormal body composition and thus be unreliable, (2) uncertainty as to whether the respiratory quotient can be used to judge the adequacy of the caloric intake, etc. The problems of extrapolation are central to using calorimetry to guide nutrition support. McClave and coworkers emphasize that intake is expressed in total calories per day while a single measurement of gas exchange is only for a few minutes. This represents a ‘snapshot’ and may not provide an accurate measurement of daily energy expenditure when extrapolated to 24 hours. The ultimate questions abuot energy measurements have to do with clinical relevance or patient outcome. For example, if one assumes that a given brief measurement is accurate and can be extrapolated to a 24 hour value, how large a difference can exist between the caloric intake and expenditure before it influences clinical outcome? McClave and associates have made a logical proposal that attention should be directed toward the cumulative energy balance over days and to determine how large a negative energy balance can be sustained before organ and tissue function are compromised. Any ‘allowable’ energy deficit would presumably be less in the presence of pre-existing malnutrition. Hyperglycemia and insulin resistance The reports of Cuthbertson [4] in the 1930s on the metabolic response to injury emphasized the urinary loss of nitrogen and other intracellular constituents, which occurred in the post-shock or flow phase of convalescence. During the years of World War II, Cuthbertson [5] stressed the difference between the ‘ebb’ or shock phase and the subsequent ‘flow’ or catabolic phase. The hyperglycemia associated with injury did not receive attention until the 1950s as the ‘diabetes of injury’. This was attributed to the anti-insulin effects of one or another of the stress hormones which led in the 1970s to the belief that it was the result of the combination of catecholamine, cortisol and glucagon stimulation [6]. The inflammatory cytokines and their neuroendocrine relationships to the hyperglycemia of stress states have been reviewed by Welborn and Moldawer [7], with attention to the ebb versus flow phases. Some of the most interesting investigations into the changes in glucose and insulin metabolism after injury have come from Ljungqvist and his group in Stockholm. The paper by Thorell et al. (pp. 69-78) in this issue reviews their studies on the development of insulin resistance. They make the broad statement that ‘the metabolic response to injury can, at least theoretically, be explained by a loss or reduction of the anabolic action of insulin’. These authors have shown that the sensitivity, as demonstrated by the hyperinsulinemic clamp technique, shows considerable variation in the normal non-diabetic patient. However, the relative reduction after a given operation is quite reproducible. With moderate operative trauma, the reduction occurs by the first postoperative day, remains for approximately 5 days and requires up to 3 weeks to return to normal. The reduction shows a dose-response relationship to the degree of trauma, yet not to age, body size, length of operation or preoperative insulin sensitivity. The reduction in insulin sensitivity on the first postoperative day was significantly correlated with the length of hospital stay. The role of glucose transporters such as GLUT4 may be involved in the altered glucose metabolism of operative trauma, since these transporters must move from their intracellular location to the plasma membrane to contribute to glucose uptake. In the preoperative control state, patients were found to show a 1.8 fold increase in GLUT4 translocation during hyperinsulinemia, while no increased occurred after operation. The combination of tracer studies with the hyperinsulinemia clamp makes it possible to separate peripheral glucose uptake from suppression of endogenous glucose production. The changes in peripheral glucose uptake were assumed to be mainly in muscle, although adipose tissue may be involved to a lesser extent. Nygren and others in this group found that normal subjects responded to immobilization and hypocaloric intake with insulin resistance in peripheral tissues. Further studies revealed that hypocaloric nutrition was more important than immobilization in producing the postoperative changes. The fed state in animals has been associated with a lessened stress hormone response, improved fluid homeostasis, improved muscle function and reduced bacterial translocation. Anesthesiologists are accepting that patients may withstand operation better if some carbohydrate is provided in the immediate preoperative period. Ljungqvist and coworkers have demonstrated that patients undergoing hip replacement could have their expected postoperative reduction in insulin sensitivity abolished by continuous infusion of insulin and glucose through the operation. This group found that 400 ml of a beverage containing 12.5% carbohydrate could be given shortly before an elective operative procedure. The beverage cleared the stomach promptly and produced a rise in plasma insulin comparable to that after a standard meal, and was associated with a reduction in the expected postoperative change in insulin sensitivity. This reduction in postoperative insulin sensitivity with preoperative glucose might be beneficial through its stimulus of insulin secretion, producing the anti-catabolic effects of insulin such as demonstrated by Allison and coworkers [8] in the 1970s. The implications are far-reaching for standard surgical practice as well as for patient care in the intensive care unit. Relation of hyperglycemia to infection Hyperglycemia may be the consequence of the systemic inflammatory response serving as a marker of the severity of stress and the degree of immunocompetence associated with infection or injury. Khaodhiar and associaties (pp. 79-82) have reviewed emerging evidence that hyperglycemia in the perioperative period can also be a significant risk factor for development of nosocomial infection. They note that counter-regulatory hormones produce both central insulin resistance with increased hepatic glucose output, and peripheral insulin resistance with suppressed uptake of glucose. The diabetic patient is unable to respond to the hyperglycemia with adequate insulin secretion. The inflammatory cytokines (tumor necrosis factor, interleukin-1 and interleukin-6) stimulate secretion of the stress hormones and also appear to produce insulin resistance directly in liver and muscle. In patients with poorly controlled diabetes, abnormalities in white cell function such as granulocyte adherence, chemotaxis, and phagocytosis all improve with glycemia control. Superoxide radical production in polymorphonuclear leukocytes is decreased in both poorly controlled diabetics and healthy non-diabetic controls that have been exposed to high concentrations of glucose in vitro. A strong correlation has been reported between the incidence of infection and the mean plasma glucose level. The authors analyze a study of 93 diabetic patients undergoing major cardiac or abdominal surgery, that were found to have a correlation between the extent of hyperglycemia on the first postoperative day and subsequent development of infection [9]. Increased infection subsequently occurred when the blood glucose exceeded 220 mg/dl, a level above which poor granulcoyte function has been shown. There was no correlation with the blood sugar on the second postoperative day, suggesting to the authors that the period of highest risk of subsequent infection occurs early after operation since that is the time of maximum stress and immunocompromise. Supporting data has been reported from patients after cardiac surgery who had a plasma glucose value over 300 mg/dl in the early postoperative period [10]. The most direct evidence that treating hyperglycemia can lower postoperative infection risk came from a study of 8 910 patients (18% of whom were diabetic) by Zarr [11]. After a 3-year experience, a special protocol of continuous insulin infusion was started to maintain the blood glucose below 200 mg/dl. The rate of postoperative infection was reduced from 2.1% to 0.98%. In a Swedish study of 630 patients with acute myocardial infarctions, the patients were randomized to very tight control using insulin infusions or usual therapy [12]. Implementation of euglycemia caused a prompt and sustained improvement in all-cause mortality. The mortality in diabetic patients was reduced by 30% in one year and 11% in long-term mean follow-up of 3.5 years. Khaodhiar and associates consider that the healing of a myocardial infarct may be analagous to healing a surgical wound and that hyperglycemia hinders wound healing (an important component of immune function). However they also point out that glucose homeostasis may have an influence on many other aspects of immune function including non-specific glycation of complement or immunoglobulin. The authors summarize the evidence that it is the timing of hyperglycemia that provides the best clue as to whether this sign is a ‘harbinger of underlying infection or a potential risk factor for future infection’. Improved glycemia control is beneficial no matter whether an infection is present or may subsequently develop. The gut and immune function The review by Stig Bengmark (pp. 83-85) on gut microenvironment and immune function is typically provocative and presents yet another challenge to conventional thinking. The concept of disease arising from a defective intestinal barrier which allows the passage of bacteria or toxic products into the systemic circulation was proposed by Metchnikoff eighty years ago, again by Cannon in 1923 and subsequently by Fine in 1959 [13]. This idea lost credence when it could not be supported by studies in germ-free animals. However, it has received new attention during the past decade, perhaps coinciding with recurrent interest in the effect of endotoxin on cytokine production. The paper by Bengmark emphasizes that the dominant interest in bacterial translocation during the past decade, has resulted in neglecting the effects of stress and malnutrition on decreasing the immune function of the gut associated lymphoid system itself. Central to this argument is the role of inflammatory cytokines which he believes are central to pathology of the bowel wall. Wanke and Bistrian [10] have reported that TNF and IL-1 can increase bacterial adherence to intestinal mucosa. There is evidence that intestinal ischemia may participate in a septic response by priming inflammatory cells capable of injuring other organs such as the lung [14]. This has led to the suggestion that cytokine assays should be used to monitor patients and identify those at risk for sepsis and other complications. The GI tract contains an extensive immune system with a total number of immune cells approaching the number found in the entire rest of the body. This gut associated lymphoid tissue (GALT) is separate and distinct from the systemic immune system. Bengmark notes a possible advantage of the GALT being separate from the ‘bone marrow’ system. In the future, the transplant patient may have the marrow-related immune responses suppressed to avoid rejection, while enhancing the gut lymphatic tissue and reducing the septic complications of immunosuppression. Lymphocyte traffic assumes importance because mature lymphocytes recirculate continuously between the blood and lymphoid organs. Malnutrition alters the return of lymphocytes from within the mesenteric lymph nodes to the intestine. An alteration in GALT has been reported with protein-calorie malnutrition, vitamin A deficiency and glutamine deprivation [15]. As he points out, we may have reached the limits of the benefits to be obtained from antibiotics, which some cases may be counterproductive through damage to the probiotic intestinal flora. The possibility that the infectious complications of surgical operation may be reduced by the therapeutic administration of probiotic bacteria, such as Lactobacillus plantanum, and of nutrients which foster their growth is a fascinating one. He also considers the idea that the immune system may be treated differently in patients undergoing organ transplantation. It may be possible, for example, to suppress the marrow-relatd immune responses, while at the same time enhancing gut-associated lymphatic tissue and reducing the septic complications of immunosuppression. Much research has focused on the effect of nutrients upon the small bowel mucosa, but, as Bengmark points out, a dominant part of the immune system and the largest bulk of bacteria are in the large intestine. Immunomodulatory treatment should not only be introduced within hours of the onset of acute illness or injury in order to modify the early inflammatory response, but this treatment also needs to be addressed to the functional requirements of the large bowel and its associated lymphoid tissue. The acute phase response, including insulin resistance, is activated quickly (Cuthbertson's ebb phase) and therefore must be treated promptly while avoiding therapy which inadvertantly depresses GALT. He urges that enteral nutrition be provided early with avoidance of saturated fats which depress immunity, avoidance of broad spectrum antibiotics, and provision of fiber which would insure nutrients for the essential gut flora. Nutritional support in the elderly Professor Vellas (pp. 87-92) from the distinguished group at Toulouse, led by Professor Alberade, has reviewed nutritional support in the elderly from a broader perspective than just the postoperative period. He points out that nutritional support is unlikely to be effective unless a diagnosis of malnutrition is first made, and this requires adequate routine protocols for initial screening and subsequent in-depth assessment of elderly patients. This is particularly important since it is among the malnourished elderly that the most striking benefits of nutritional support are found. He also points out that it is among elderly patients that malnutrition is most prevalent and that, in the Western world at least, this group is getting ever larger. He proposes a nutritional scoring system for the elderly, validated by his own group, which has been used to study the epidemiology of malnutrition in a French community and hospital population. He also describes studies which have an important bearing on the nutritional requirements of the elderly, and in which he showed that elderly women with energy intakes at or below the RDA of 25 kcal/kg, were more likely to become frail, sick or die within 10 years than those with energy intakes in the mid-range. The paper by this group also showed that women with protein intakes of 1.2 to 1.6 g/kg/day in 1980 and 1981 tended to have fewer health problems over the next 10 years than those with protein intakes less than 0.8 g/kg/day. This presents a challenge to current thinking and has important implications for the care of hospital patients as well. The growing number of studies of nutritional support in the elderly are reviewed, but this is clearly an area where much further work needs to be done. Care of the elderly is one of the largest medical challenges in the Western world and poses an ever increasing demand upon health care resources. Simple and cost effective nutritional strategies are needed and the evidence base for these strategies must be further developed. The papers presented in this section represent an interesting commentary on the status of nutrition support approximately thirty years after its introduction. Our understanding of the metabolic response to injury and illness remains a work in progress. Yet these emerging changes in our concepts suggest a future where nutritional therapy can offer much more selective and well-defined advantages for the convalescent patient.
No takes yet. Share an insight, caveat, or question.
Allison et al. (1999) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: