Previous - Next
R.J. MALIAKKAL, D. DRISCOLL and B.R. BISTRIAN*
* Laboratory of Nutrition/Infection, Cancer Research Institute, New England Deaconess Hospital, 194 Pilgrim Road, Boston, MA 02215, U.S.A.
Abstract
1. Introduction
2. Metabolic response to starvation
3. Metabolic response to stress
References
Critically ill patients
often have unique nutritional needs that require a rethinking of
classic nutrition teaching. Semistarvation in the unstressed,
malnourished patient is physiologically distinct from the
response in the stressed, malnourished patient.
With starvation, net protein loss proceeds from most organs, skeletal muscle being the major contributor, whereas with stress, loss from skeletal muscle is increased, while certain organs, such as liver, and the wound experience net anabolism. Adequate feeding of the semi-starved, recuperating patient will result in effective lean tissue repletion, and overfeeding to a certain extent will increase the rate of repletion without adverse consequence. Meeting or modestly exceeding caloric balance while providing an ample (1.5 g/kg) protein intake will serve only to maintain the stressed patient, whereas substantial overfeeding can be harmful to organ function.
Standard
formulations must often be modified in order to accommodate fluid
restriction, organ failure, and limited central venous access.
Malnutrition is now
generally acknowledged as a common finding in hospitalized
patients. In medical and surgical patients, the prevalence is
similar, occurring in up to 50% of hospital admissions (BISTRIAN
et al., 1974; 1976) and exceeding 25% in most acute
care hospitals. However, qualitative differences were initially
noted between patient groups with respect to type of
malnutrition. Surgical patients were generally more severely
protein-depleted, with both a reduced arm muscle circumference
and hypoalbuminemia. In contrast, medical patients were
principally deficient in calories, as manifested by a reduced
weight/height ratio and triceps skinfold. These findings were
thought to be related to the higher levels of metabolic stress
characteristic of surgical illness.
In the years subsequent to this rediscovery of protein-calorie malnutrition (PCM), it has become obvious that what was broadly considered PCM actually resulted from two distinct processes: semistarvation and metabolic stress (i.e., the metabolic response to injury, infection and inflammation). Although there is considerable overlap, since metabolic stress and semistarvation so often coexist, the consequence of a metabolic response to stress, prolonged beyond 10 to 14 days without adequate feeding, is a condition that has come to be known as hypoalbuminemic malnutrition.
Semistarvation for a number of weeks or months, depending on the degree of caloric deficit, produces true protein-energy malnutrition. This is true in the sense that both protein and energy are deficient to about the same degree, and adequate feeding will completely reverse it. It is termed cachexia or marasmus. It has to be distinguished from hypoalbuminemic malnutrition in which protein is more limiting than energy and which has more severe consequences. Not only is organ function impaired, but also a patient with hypoalbuminemic malnutrition is incapable of appropriately handling further metabolic stress, principally due to a limited capacity to produce interleukin 1 (IL-1), the essential trigger for the stress response released by phagocytizing mononuclear cells (HOFFMAN-GOETZ et al., 1981).
Two key points deserve strong emphasis. The metabolic response to stress is generally a net benefit to the well-nourished host in terms of wound healing, immune response, and survival for a period of 7 to 14 days without feeding and for an indefinite period with feeding. Only with an overwhelming response to endotoxin, so-called septic shock, does this response become a net detriment to the host, and even here through exaggeration of usually beneficial actions, for instance, overproduction of oxygen radicals, excessive coagulation, splanchnic vasoconstriction, and vascular permeability, rather than through some unique effects.
A corollary
is that patients with pre-existing PCM of the marasmus variety
have a much shorter grace period, and that patients with
hypoalbuminemic malnutrition have no grace period at all and can
even benefit from feeding for 7 to 10 days prior to planned
stress interventions (MULLER et al., 1982). Given this
background it becomes essential for medical professionals to
recognize the magnitude and ubiquity of this problem and to
address these issues whenever providing hospital care. A thorough
understanding of intermediary metabolism, as it relates to the
metabolic response to starvation and stress and interaction with
feeding, is necessary in order to provide optimal nutritional
support to the hospitalized patient.
The decline in plasma
glucose that follows a decrease in dietary carbohydrate leads to
a fall in plasma insulin and a rise in glucagon. This fall in
insulin/glucagon ratio results in glycogenolysis, reduction of
lipogenesis, and mobilization of amino acids.
Glycogenolysis occurs via stimulation of liver glycogen phosphorylase, thereby providing energy to glucose-dependent tissue like the brain, bone marrow, erythrocytes and renal medulla. Liver glycogen stores are fully depleted after 72 hours of fasting, and gluconeogenesis, from amino acids principally, becomes the major source of new glucose formation by the second day of total starvation.
De novo lipogenesis by the liver is reduced, lipolysis increased, and free fatty acids are released from adipose tissue. Furthermore, the reduction in fatty acid synthesis results in decreased levels of malonyl coenzyme A, which is the first committed step in fatty acid synthesis and a negative feedback inhibitor of carnitine acyltransferase 1. This mitochondrial membrane enzyme is an essential component for translocating free fatty acids into the mitochondria for oxidation. If insulin levels are sufficiently low (i.e., less than 100 g of dietary carbohydrate are ingested per day), ketone body synthesis will also occur. Both free fatty acids and ketone bodies become important sources of energy for the semistarved host, and this physiologic adaptation to alternative fuel is an important survival mechanism (CAHILL, 1970).
Amino acids are mobilized from skeletal muscle, where nearly 75% of actively metabolizing tissue is found, to support protein synthesis in the other vital organs. Thus, the visceral organs (brain, heart, liver, lungs, kidneys) are to a greater or lesser extent preserved at the expense of skeletal muscle and connective tissue.
In essence,
during simple starvation in the absence of stress, the body
provides for its energy and protein needs through metabolic
adjustments and use of adipose tissue and skeletal muscle, with
relative preservation of more vital organ function and serum
albumin level. Certainly, if weight loss is severe, in the order
of 15-20%, even though the stress response can still be mounted,
the period before nutritional support becomes essential is
shortened. With even more severe degrees of weight loss, the
response to stress can be impaired through poor respiratory
muscle function and wound-healing ability, consequent to loss of
muscle and connective tissue. However, even with very severe
marasmus (30% weight loss) there is little evidence for serious
impairment of immune function.
The metabolic response to
stress has a vastly different set of priorities. Here the
mobilization of body protein, fat and carbohydrate stores is an
active process, as would be helpful in a response that requires
an increase in protein synthetic rate in certain tissues. Rather
than the passive response seen in starvation consequent to a fall
in serum insulin, mobilizing hormones, cortisol, growth hormone,
catecholamines and glucagon serve to release amino acids, glucose
and free fatty acids. Energy expenditure is increased, not
diminished as in starvation.
The initial phase of critical illness is accompanied by the diffuse discharge of autonomic mediators. The resultant release of catecholamines from the adrenal medulla and the central nervous system via postganglionic terminals produces a wide array of effects on various organ systems. The effects on the cardiovascular system are to stabilize hemodynamics and maintain vital organ perfusion. The respiratory rate and tidal volume increase to maximize oxygenation and compensate for acid-base disturbances.
This early phase response to injury peaks within 24 hours and is believed to be a consistent response irrespective of the degree of stress (DOUGLAS and SHAW, 1989). This neurohormonal response is centrally mediated through the release of hypothalamic and pituitary hormones, including ACTH, corticotrophin-releasing factor and antidiuretic hormone. A second major component in the metabolic response to stress is the release of IL-1 and tumor necrosis factor (TNF) from phagocytosing tissue macrophages and circulating monocytes.
These cytokines share certain activities such as the development of fever, white blood cell activation and release, acute phase protein production, and a fall in serum albumin, iron and zinc. IL-1 uniquely underlies the T-cell and B-cell response of cellular immunity and antibody production, the increase in energy expenditure, and the changes in glucose metabolism including increases in glucose flux and oxidation, as well as insulin, glucagon, catecholamine, and ACTH secretion. TNF increases skeletal protein catabolism and, in large doses, mimics endotoxic shock (for a fuller discussion of all these changes, see SCHWARTZ and BISTRIAN, 1990; POMPOSELLI, FLORES and BISTRIAN, 1988; DINARELLO, 1986). IL-1 and TNF also have additive or synergistic effects on many of the shared as well as the individual actions. The hallmark of the stress response is the precipitous drop in serum albumin concentration (MOSHAGE et al., 1982; RICKETTS and BULL, 1962; FLECK et al., 1985).
Just as there are differences in pathophysiology and clinical presentation of marasmus and hypoalbuminemic malnutrition, there are also profound differences in response to nutritional support and important variants in nutritional requirements. Meeting protein and energy goals in the marasmic patient will lead to efficient regain of lean tissue at about the same rate at which it was lost. Overfeeding to a reasonable degree will accelerate this process (SMITH et al., 1977; SHAW et al., 1983), whereas repletion is not possible in the hypoalbuminemic patient with ongoing stress (STREAT, BEDDOE and HILL, 1987).
Nutritional needs too are quite different. Energy requirements reflect total energy expenditure which is composed of the resting metabolic rate (RMR), activity energy expenditure, and the thermal effect of food. Since indirect calorimetry is not widely available, the RMR is usually estimated by the Harris-Benedict formula which gives values of approximately 22-25 kcal/kg/d.
Repletion is the goal in the marasmic and in the seriously malnourished, and energy intakes modestly above expenditure are not a risk after initial stabilization. Providing calories at 35-40 kcal/kg/d (or about 1.5 RMR) will meet total energy expenditure as well as provide for the energy cost of tissue repletion of about 4 kcal/g. The maximal repletion rate is 4-6 g nitrogen per day, representing 25 to 37.5 g protein or 120 to 180 g lean tissue, with an additional 10% as adipose tissue, in the usual 50/50 caloric ratio of protein to fat repletion. This represents a maximal rate of weight gain of 130 to 200 g/d.
Weight gain in excess of this represents fluid excess, due principally to the antinatriuretic and antidiuretic effects of insulin evoked by the usual high-calorie, high-carbohydrate feeding regimes. This weight gain can and should be avoided by limiting sodium and fluid intake and using a mixed fuel regimen of fat and carbohydrate.
For the patient with hypoalbuminemic malnutrition, energy needs range from RMR to substantially above RMR, depending on the important variables, age and severity of stress. If the patient is a young adult and has severe stress (burns, multiple trauma, head injury, severe sepsis), energy requirements are in the range of 1.3 to 1.7 RMR. However, the far more commonly seen patient who requires parenteral nutrition is older, post-surgical, and often artificially ventilated. In such patients, the total energy expenditure and energy goal is 1 to 1.2 RMR (HUNTER et al., 1988; ASKANAZI et al., 1980a; ROULET et al., 1983).
Carbohydrates are the principal energy source in total parenteral nutrition. The liver produces glucose at 2 mg/kg/min in the postabsorptive state (200 g/d in the 70 kg man), but this production is completely suppressed by total parenteral feedings in normal individuals and in marasmic patients.
In the patient with an active stress response, there is little suppression of endogenous glucose production, particularly with parenteral glucose, where only 20% of the cardiac output is directed to the liver, the main site of gluconeogenesis. Thus, total glucose flux may be almost doubled which, in conjunction with the stress hormones (catecholamines, cortisol, growth hormone, cytokines) producing insulin resistance, impairs glucose clearance leading to hyperglycemia.
Impaired glucose tolerance, unmasked by parenteral nutrition in the stressed patient, can lead to an increased risk for nosocomial infection if blood glucose levels exceed 200 mg/dL (OVERETT et al., 1986; BAXTER et al., 1990). Since lipogenesis in the liver is a less insulin-resistant activity (RYAN, BLACKBURN and CLOWES, 1974) and cytokines increase hepatic lipogenesis (GRUNFELD, DINARELLO and FEINGOLD, 1991), fatty liver can result. Thus, when feeding the acutely ill, 200 g of dextrose can be provided over the first 24 to 48 hours. Further increments in dextrose can be made slowly to meet caloric goals, provided blood glucose control is being achieved.
The upper limit of dextrose administration should not exceed 4 mg/kg/min (WOLFE et al., 1980). Feeding above this optimal infusion rate increases the non-oxidative disposal by glycogen and lipid formation. As glycogen stores are limited and rapidly filled, hepatic lipogenesis predominates, leading to excessive production of carbon dioxide as a consequence of the elevated respiratory quotient of de novo lipogenesis. This can interfere with weaning from artifical ventilation in patients with limited reserves (ASKANAZI et al., 1980b).
Although less well documented in man than in animals, stressed patients do not tolerate parenteral fat in large quantities (SEIDNER et al., 1989). Preliminary results suggest that cyclic administration of as little as 1000 fat calories/day may significantly increase infection rate (MULLER et al., 1986; BUZBY et al., 1991). Lipids can be used to provide a part of the non-protein calories, usually up to 30%.
The incorporation of lipids into the total parenteral formula, forming a three-in-one or total nutrient admixture, has made administration of lipids safer and also simpler. However, in critically ill patients, especially those who have excessive total body water from over-administration of intravenous fluids, lipids should not be employed until parenteral nutrient volumes approach 1500 cm3/d.
Glucose is a preferred fuel for metabolic as well as protein-sparing effects until 2-3 mg/kg/min are provided. Essential fatty acid deficiency in the plasma will not ensue, since ample stores of linoleic acid in the adipose tissue (about 10%) are generally present in most patients and will be released whenever the feeding regimen is hypocaloric.
The lipid nutrients available in parenteral form are long-chain triglycerides (LCT) derived from safflower and/or soybean oil. These provide the full complement of essential fatty acids. The LCTs are degraded peripherally by lipoprotein lipase. However, there is also uptake by the reticuloendothelial system which can interfere with RES function when the lipid load is large (SEIDNER et al., 1989). Provision of lipids in smaller amounts, via a 24-hour continuous infusion as a total nutrient admixture, does not interfere with RES function (JENSEN et al., 1990).
New lipid formulations of physical mixtures of medium-chain triglycerides and LCT, available at present in Europe and being investigated in the United States, appear to have superior energetic properties and do not interfere with the RES (SEIDNER et al., 1989). On the horizon are structured lipids derived by hydrolysis, and then random re-esterification of LCTs and MCTs that have unique anticatabolic effects and maintain RES function (MASCIOLI et al., 1987).
A new therapeutic possibility that has excited clinicians and nutritionists is the use of special foods for medical purposes. Although nutrients such as glutamine, arginine, nucleotides, structured lipids, and omega-3-oils when provided enterally need not undergo evaluation as drugs, it is appropriate that their parenteral administration require a greater degree of scrutiny. Particularly with omega-3-fats we have entered an era where beneficial modulation by nutrition of the excessive response to stress, as exemplified by septic shock or severe burns, may be possible (POMPOSELLI et al., 1991; GOTTSCHLICH et al., 1990; DALY et al., 1991).
Protein is the final energy-containing nutrient that merits discussion in relation to the differences between marasmic and hypoalbuminemic malnutrition. For the well-nourished individual without stress, the Recommended Dietary Allowance is 0.8 g protein/kg/d. For the marasmic patient, provision of protein at 1.5 g/kg/d will provide more than adequate amounts of protein for repletion, but will not tax renal or hepatic function except with severe renal (creatinine clearance less than 20 cm3/min) or hepatic (Childs B or C) dysfunction.
For the patient with hypoalbuminemic malnutrition, protein needs to achieve balance are greater secondary to the catabolic response which diminishes the efficiency of protein utilization. The usual recommendation is 1.5 to 2 g/kg of high-biologic-value protein, when hepatic and renal function are adequate. Protein in excess of this amount leads only to ureagenesis, even in the most severely stressed (WOLFE, GOODENOUGH and BURKE, 1983).
Branched-chain amino acid (BCAA) enriched parenteral nutrition which provides up to 50% of the amino acids as leucine, isoleucine and valine, rather than the more usual less than 20%, seems to have a clinical as opposed to a biochemical advantage in two settings. In severe hepatic failure, general consensus is that BCAAs are better tolerated and relieve hepatic encephalopathy (NAYLOR et al., 1989). The improved protein utilization also gives better nitrogen balance for a given degree of protein intake in renal impairment (ECHENIQUE et al., 1984). BCAAs should not be used in amounts of less than 40 g/d, because other essential amino acids become limiting at this level of protein intake. Other special amino acid formulations, such as low doses of essential amino acid only (i.e., less than 20 g/d), have fallen into disfavor, since, although this amount and formulation will minimize urea production, it is grossly inadequate in terms of net protein economy.
There are a number of other consequences of the injury response that must be considered when providing parenteral nutrition. The antidiuretic and antinatriuretic effect of the combination of aldosterone, ADH, and insulin secretion, resulting from the stress response, make the stressed patient intolerant of large fluid intakes. However, post-injury patients require large quantities of fluid to maintain intravascular volume, given the increased permeability of small vessels with stress. Finally, many medications including pressors, antifungal agents and antibiotics, necessary for clinical care, require substantial quantities of fluid for their administration.
These three factors often combine to make volume requirements the limiting factor in the quantity of nutrition that can be given. With amounts often limited to 1000 cm3/d, it has been our practice to maximally concentrate parenteral nutrition in one liter containing 7% amino acids and 21% dextrose. This provides 70 g protein and 210 g dextrose which appears to be a reasonable compromise between protein and carbohydrate requirements, i.e., 1 g/kg/d and 2 mg/kg/min, respectively, for the 70 kg reference man.
Secondly, the normal production of acid in man is approximately 1 mEq/kg/d, but this may double or even triple as a result of increased protein catabolism. Acid excretion by the kidney is often impaired by drug- or disease-induced dysfunction. Loss of small intestinal secretion through secretory diarrhea, ileostomy losses or biliary secretions are bicarbonate-rich. All of these conditions lead to metabolic acidosis. On the other hand, loss of hydrogen ion from nasogastric suction, or chloride imbalance that develops with diuresis, leads to metabolic alkalosis.
Since protein utilization and many other cellular processes work best at a physiologic pH, the parenteral feeding formula can often be the vehicle for managing acid-base homeostasis, employing sodium and potassium acetate to manage metabolic acidosis, and hydrochloric acid for metabolic alkalosis.
One final example is the hypercoagulable state that is a consequence of metabolic stress. Since parenteral nutrition is provided through a central venous catheter to allow dilution of the hypertonic admixture, thrombosis is an acknowledged severe complication of total parenteral nutrition. To limit this occurrence, inclusion of 6000 units of heparin/d in the nutrient admixture has been demonstrated to reduce, but not eliminate, catheter-related thrombosis (IMPERIAL et al., 1983).
Certain conditions, such as inflammatory bowel disease, further increase the propensity for hypercoagulability, as reflected in a partial thromboplastin time substantially shorter than normal (ELSEN and BISTRIAN, 1991). In such individuals, provision of up to 12000 units of heparin in the parenteral nutrient formula can be employed to extend PTT toward or even to normal. One caveat is that 6000 units of heparin in the parenteral formula also serve as the equivalent to mini-dose heparin given subcutaneously for deep vein thrombosis prophylaxis, and the latter should be discontinued when heparin is furnished with the parenteral formula.
In
conclusion, hypoalbuminemic malnutrition, and more specifically
the metabolic response to stress, demands individualization of
nutritional support regimens. The goal in such patients is to
support the response rather than to attempt protein repletion.
The standard principles of protein and energy metabolism,
applicable when feeding the marasmic patient, generally either do
not apply or require extensive modification in the critically ill
patient. However, such attention has been and is worthwhile,
since metabolic support of the critically ill has changed medical
and surgical practice to a substantial degree and led to
significant improvement in clinical outcome.
ASKANAZI, J., CARPENTIER.
Y.A., ELWYN, D. et al.: Influence of total parenteral
nutrition on fuel utilization in injury and sepsis. Ann. Surg., 191,
40-46 (1980a).
ASKANAZI, J., ROSENBAUM, S.H., HYMAN, A.I. et al., Respiratory changes induced by large glucose loads of total parenteral nutrition. JAMA, 243, 1444-1447 (1980b).
BAXTER, J.K., BABINEAU, T.J., APOVIAN, C.M. et al., Perioperative glucose control predicts increased nosocomial infection in diabetics. Crit. Care Med. (Abstr.), 18, 5207 (1990).
BISTRIAN. B.R., BLACKBURN, G.L., HALLOWAL, E., MEDDLE, R.: Protein status of general surgical patients. JAMA, 230, 858-860 (1974).
BISTRIAN, B.R., BLACKBURN, G.L., VITALE, J. et al., Prevalence of malnutrition in general medical patients. JAMA, 235, 1567-1570 (1976).
BUZBY, G. et al., Perioperative total parenteral nutrition in surgical nutrition. N. Engl. J. Med., 325, 525-532 (1991).
CAHILL, G.F.: Starvation in man. N. Engl. J. Med., 282, 668-675 (1970).
DALY, J.M., LIEBERMAN, M., GOLDFINE, M.S. et al.: Enteral nutrition with supplemental arginine, RNA and omega-3 fatty acids: A prospective clinical trial. JPEN (Abstr.), 15, 17 (1991).
DINARELLO, C.A.: Interleukin-1: Amino acid sequences, multiple biological activities and comparison with tumor necrosis factor (cachectin). Year Immunol., 2, 68-89 (1986).
DOUGLAS, R.G., SHAW, J.F.: Metabolic response to sepsis and trauma. Br. J. Surg., 76, 115-122 (1989).
ECHENIQUE, M.M., BISTRIAN, B.R., MOLDAWER, L.L. et al.: Improvement in amino acid use in the critically ill patient with parenteral formula enriched with branched chain amino acids. Surg. Gyn. Obstet., 159, 3-41 (1984).
ELSEN, R.J., BISTRIAN. B.R.: Shortened partial thromboplastin time in patients with inflammatory bowel disease requiring hyperalimentation. Gastroenterology (Abstr.), 100, A521 (1991).
FLECK, A., RAINES, G., HAWKER, T. et al.: Increased vascular permeability: a major cause of hypoalbuminemia in disease and injury. Lancet, 1, 781-784 (1985).
GOTTSCHLICH, M., JENKINS, M., WARDEN, G.D. et al.: Differential effects of three enteral dietary regimens on selected outcome variables in burn patients. JPEN, 14, 225-236 (1990).
GRUNFELD, C., DINARELLO, C.A., FEINGOLD, K.R.: Tumor necrosis factor-alpha, Interleukin-1, and Interferon alpha stimulate triglyceride synthesis in HepG2 cells. Metabolism, 9, 894-898 (1991).
HOFFMAN-GOETZ, L., McFARLANE, D., BISTRIAN, B.R., BLACKBURN, G.L.: Attenuation of fever and relative hyperferremia in rabbits infected with endogenous pyrogen harvested from blood leukocytes of malnourished patients. Am. J. Clin. Nutr., 34, 1109-1116 (1981).
HUNTER, D.C., JAKSIC, T., LEWIS, E. et al.: Resting expenditure in the critically ill: Estimations versus measurements. Br. I Surg., 75, 875-878 (1988).
IMPERIAL, J., BISTRIAN, B.R., BOTHE, A. et al.: Limitation of central vein thrombosis in total parenteral nutrition by continuous infusion of low-dose heparin. J. Am. Coll. Nutr., 2, 63-73 (1983).
JENSEN, G.L., MASCIOLI, E.A., SEIDNER, D.L. et al.: Parenteral infusion of long and medium chain triglycerides and reticuloendothelial system function in man. JPEN, 6, 383-388 (1990).
MASCIOLI, E.A., BISTRIAN. B.R., BABAYAN, V.K. et al.: Medium chain triglycerides and structured lipids as unique non-glucose energy sources in hyperalimentation. Lipids, 22, 421-423 (1987).
MOSHAGE, H.J., JANSSEN, J.A.M., FRANSSEN, J.H. et al.: Study of the molecular mechanism of decreased liver synthesis of albumin in inflammation. J. Clin. Invest., 29,1635-1641 (1982).
MULLER, J.M., DIENST, C., BRENNER, E. et al., Preoperative parenteral feeding in patients with gastrointestinal carcinoma. Lancet, 1, 68-71 (1982).
MULLER, J.M., KELLER, H.W., BRENNER, E. et al., Indications and effects of preoperative parenteral nutrition. World J. Surg., 10, 53-63 (1986).
NAYLOR, C.D., O'ROURKE, K., DETSKY, A., BAKER, J.P.: Parenteral nutrition with branched chain amino acids in hepatic encephalopathy. Gastroenterology, 97, 1033-1042 (1989).
OVERETT, T., BISTRIAN, B.R., LOWRY, S. et al., Total parenteral nutrition in patients with insulin-requiring diabetes. J. Am. Coll. Nutr., 5, 79-89 (1986).
POMPOSELLI, J.J., FLORES, E.A., BISTRIAN, B.R.: Role of biochemical mediators in clinical nutrition and surgical metabolism. JPEN, 12, 212-218 (1988).
POMPOSELLI, J.J., FLORES, E.A., BLACKBURN, G.L. et al., Diets enriched with N-3 fatty acid ameliorate lactic acidosis by improving endotoxin-induced tissue hypoperfusion in guinea pigs. Ann. Surg., 213, 166-176 (1991).
RICKETTS, C.R., BULL, J.P.: Studies of plasma protein metabolism. Clin. Sci., 23, 411-423 (1962).
ROULET, M., DETSKY, A., MARLISS, E.B. et al.: A controlled trial of the effect of parenteral nutritional support on patients with respiratory failure and sepsis. Clin. Nutr., 2, 97-105 (1983).
RYAN, N.T., BLACKBURN, G.L., CLOWES, G.H.A., Jr.: Differential tissue sensitivity to elevated endogenous insulin levels during experimental peritonitis in rats. Metabolism, 23, 1081-1089 (1974).
SCHWARTZ, J.H., BISTRIAN, B.R.: The role of cytokines in intermediary metabolism. In: Cellular and Molecular Aspects of Endotoxin Reactions, pp. 427-445, A. NOVOTNY, J.J. SPITZER, E.J. ZIEGLER (Eds.). Elsevier Science Publishers, Amsterdam, Netherlands, 1990.
SEIDNER, D.L., MASCIOLI, E.A., ISTFAN, N.W. et al.: Effects of long-chain triglyceride emulsions on reticuloendothelial function in humans. JPEN, 13, 614-619 (1989).
SHAW, S.N., ELWYN, D.H., ASKANASI, J. et al.,: Effects of increasing nitrogen intake on nitrogen balance and energy expenditure in nutritionally depleted patients receiving parenteral nutrition. Am. J. Clin. Nutr., 37, 930-940 (1983).
SMITH, M.F., BLACKBURN, G.L., BISTRIAN, B.R., GRIFFIN, R.E.: Beneficial effects of high nitrogen-calorie (N:cal) ratio in intravenous hyperalimentation. Surg. Forum, 28, 63-64 (1977).
STREAT, S.J., BEDDOE, A.H., HILL, G.L.: Agressive nutritional support does not prevent protein loss despite fat gain in septic intensive care patients. J. Trauma, 27, 252-266 (1987)
WOLFE, R.R., GOODENOUGH, R.D., BURKE, J.F.: Response of protein and urea kinetics in burn patients to different levels of protein intake. Ann. Surg., 197, 163-171 (1983).
WOLFE, R.R., O'DONNELL, T.F., STONE, M.D. et al.: Investigation of factors determining the optimal glucose infusion rate. Metabolism, 29, 892-900 (1980).