Towards the optimization of enteral nutrition

CLINICAL

THE ARVID WRETLIND

LECTURE

NUTRITION

(1987) 6:61-74

1986

Towards the optimization of enteral nutrition D. B. A. Silk Department of Gastroenterology

&Nutrition,

Central Middlesex Hospital, Acton Lane, London NW10 7NS, UK.

INTRODUCTION

One of the most vexing problems facing investigators has been how to monitor the efficacy of nutritional support, and in this regard enteral nutrition is no exception. Since we have found it practically and technically possible to calculate nitrogen balance on a regular basis we have used this parameter as a judge of efficacy. We would not, however, necessarily advocate that this is the best or only method available and in this regard the recent work on the effects of parenteral nutrition on muscle function is of considerable interest [8].

It is 10 years since Bistrian and colleagues published their papers [ 1, 21 showing that up to 50% of hospitalised medical and surgical patients had some evidence of nutritional deficiencies. One interpretation of their findings was that up to 50% of the patients were ‘malnourished’ and therefore in need of nutritional support. Subsequently it has become clear that it is difficult to agree on what actually represents clinically significant malnutrition [3]. Moreover, even if agreement is reached, it may not be clear that outcome will be prejudiced by providing nutritional support. This is so because in the hospitalised patient malnutrition often arises as a consequence of the underlying disease process which may often not be amenable to correction. In the light of these comments it is not surprising that it is frequently not possible to clearly define the indications for instituting nutritional support. The problems involved in reaching a decision as to whether to institute nutritional support or not have been reviewed by several authors [4,6]. Once the decision has been made to institute nutritional support, it can be provided by the enteral or parenteral route. Clinical experience during the last decade has confirmed that enteral nutrition is an efficient way of providing nutritional care to patients with normal or near normal gastrointestinal function, and two surveys carried out in our own unit during the last 7 years have shown that over three quarters of all patients receiving nutritional support have been fed enterally. It is because enteral nutrition has found such wide application that we have directed a good deal of our research efforts towards optimising its efficacy. Our overall aims have been firstly to increase our understanding of the basic physiology of nutrient assimilation in normal and disease states in order to apply principles gained to the formulation of enteral diets. In addition, clinical research has been directed towards improving the techniques involved in administering enteral nutrition to patients. The term enteral nutrition commonly refers to the infusion of a formulated enteral diet with or without the use of a peristaltic pump via per nasal fme bore feeding tubes, positioned in the stomachor small intestine [7].

Table 1 Nitrogen balance during enteral feeding

Diet -] Vivonex’b HN a 1 Clinifeedg

b

Clinifeedn 400 Clinifeed@ 400 i Expresssm EFl

Values are mean f SE Duration of feeding n (days)

Nitrogen balance g/24 h

25 21

16.0 f 1.6 20.3 f 2.8

+0.6 f 0.8 +2.6 f 0.6

30 29 27

10.9 f 8.7 9.9 f 8.9 7.9 f 5.9

-1.2 f 0.4 +0.4 f 0.3 +Z.8 f 1.1

(a) Data from Jones et al [9], (b) Data from Keohane et al [IO].

Table 1 summarises the nitrogen balance data observed in two of our controlled clinical trials [Q, lo]. It can be seen that the results are far from impressive and up to 400/b of individual patients entered in to the two trials remained in overall negative nitrogen balance despite continuous 24 h enteral feeding. It is now clear that in traumatised and septic hypermetabolic patients even on adequate intake of nitrogen and calories may not overcome the primary neuroendocrinological responses [l l] and at least in the early phases these patients will remain in overall negative nitrogen balance. In other patients, however, negative nitrogen balance is more likely to persist on account of inudequate intakes.

We have identified a number of factors that are likely to result in inadequate nutritional intake in enterally fed patients. Summarised in Table 2, these include the choice of the wrong type of enteral diet, the use of an inadequately formulated diet, the poor performance of 61

TOWARDS THE OPTIMIZATION

62

OF ENTERAL NUTRITION

Table 2 Factors leading to inadequate during enteral nutrition 1. 2. 3. 4. 5. 6. 7.

nutritional

intake

Incorrect choice of enteral diet Inappropriate diet formulation Poor performance of enteral feeding tubes Use of small volume diet containers Inappropriate route of administration Routine use of ‘starter regimens’ Slowing of infusion rates to counteract gastrointestinal side effects and aspiration

enteral feeding tubes, the use of small volume diet containers, the wrong route of administration, and the routine use of ‘starter’ regimes. The ensuing article discusses some of our experiences which have led us to believe that there are simple methods available whereby nutritional intake can be increased thereby improving the efficacy of enteral nutrition.

Enteral diets Table 3 summarises the different categories of enteral diets available. Until recently controversy has existed as to whether enteral feeds formulated and prepared in the hospital dietetic department should be used in preference to their commercially prepared counterparts. Problems with infection have now been well documented with ‘home brew’ diets [12] and there is now controlled data to show that the incidence of diarrhoea is higher when ‘home brew’ rather than commercial diets are used for enteral feeding [ 131. We therefore still recommend the routine use of the commercially available diets. Knowledge gained about the processes involved in the physiology of nutrient absorption and subsequent metabolism both in normal and disease states continues to influence the formulation of enteral diets, and recent advances in these areas that are influencing diet formulation are discussed.

Fat absorption There has been little research that has actually influenced diet formulation in the last 2 years. The importance of preventing essential fatty acid deficiency needs emphasising, particularly since it may occur in situations where gut function is particularly impaired and long chain triglycerides are poorly assimilated [ 141. This author continues to point out [15] that medium chain triglycerides (MCTs) are probably not as well assimilated in the presence of bile salt deficiency, exocrine pancreatic insufficiency and intestinal resection as many would believe [16] and at least in the latter two situations pancreatic supplements will greatly enhance energy assimilation from MCTs and reduce steatorrhoea [ 171.

Carbohydrate The physiology of digestion and absorption of dietary carbohydrate in relationship to formulation of enteral diets has been reviewed [ 181. Although the early chemically defined ‘elemental’ diets administered to man contained glucose as the sole carbohydrate energy source [19, 201, the glucose content was soon replaced in part by sucrose [21], and later by glucose polymer mixtures derived from the hydrolysis of starch by cc-amylase [ 181. Although the potential carbohydrate forms used in enteral diets have been categorised [22], Jones and colleagues [23] have characterised in detail the composition of starch hydrolysates commonly used as the carbohydrate energy source of enteral diets. Using gel permeation chromatography they found that most consist of a very heterogenous mixture of glucose polymers approximately 50yb of the glucose content being present as polymers containing more than 10 glucose molecules, the remaining glucose content being present as shorter chain polymers containing less than 10 glucose molecules [23]. Intestinal perfusion studies per-

Table 3 Classification of enteral diets

Type

Comments

Polymeric

Protein nitrogen source. For use in patients with normal or near normal gastrointestinal function

Predigested ‘chemically defined elemental’

Free amino acid or oligopeptide nitrogen source. Small quantities of long chain triglycerides. For use in patients with severe gastrointestinal disease

‘Disease specific diets’ Portosystemic encephalopathy

Free amino acid nitrogen source. High branched chain amino acid content, low aromatic ammo acid content. Indications still under discussion [ 1031

Stress

Branched chain amino acid enriched. Indications still under discussion [90]

Renal failure

Enteral nutrition in renal failure discussed elsewhere [ 1041

Respiratory and cardiac failure

Protein nitrogen source. Reductions in carbohydrate component of energy source [ 1051

CLINICAL

formed in normal human volunteers in the absence of luminal cc -amylase activity showed a differential handling of the glucose polymers by the jejunum. The higher MW glucose polymers (containing > 10 glucose molecules) were assimilated slower than the lower MW glucose polymers. It seemed that the latter could be conferring a kinetic advantage on glucose transport [23]. This was confirmed in a further study [24] in which purified high and low MW polymer fractions were perfused. Although the oz -amylase hydrolysate of the low MW glucose polymer fraction (< 10 glucose molecules) conferred the expected kinetic advantage on glucose transport, the high MW fraction (osmolality 2 of the starting material) was surprisingly well absorbed even in the absence of x-amylase [24]. The authors concluded that the energy content of enteral diets could be increased at the same time as lowering diet osmolality by substituting the commonly used heterogenous starch hydrolysates with purified high MW fractions. To date in only one diet (Vivonex TEN”, Norwich Pharmaceuticals Inc) has this concept been utilised, with a subsequent lowering of osmolality from 830-630 mosmol/kg. Lactose Lactose deficiency is common in certain racial and ethnic groups, particularly those of African descent, Asians and Jews, with an incidence‘as high as 60-1009, [25]. On account of this it is perhaps hardly surprising that it is now considered commercially wise to minimise the lactose content of new enteral products. Walike & Walike [26] were the first to suggest that the lactose content of blenderised formulae was the cause of diarrhoea that occurred during enteral feeding. Certainly the recent study of O’Keefe and colleagues [27] demonstrated that if patients with biochemically proven lactose malabsorption are bolus fed with lactose containing enteral diets, then diarrhoea will occur with stool volumes in excess of 1 L/24 h. It must be mentioned, however, that it is far from proven that it is the lactose content that causes gastrointestinal symptoms when milk is ingested by normal subjects with biochemically proven lactose intolerance. Two studies have failed to show that the ingestion of 480 ml lactose containing milk by healthy normal subjects with biochemically proven lactose malabsorption results in a significantly higher incidence of gastrointestinal side effects as compared with when the same subjects ingested the same quantity of lactose free milk [28,29]. In the recent study of Keohane and colleagues [lo], enterally fed patients with biochemically proven lactose malabsorption did not have a higher incidence of gastrointestinal side effects when fed a lactose containing enteral diet (20.9 f 5.3 g lactose/24 h) as com-

NUTRITION

63

pared to a lactose free formulation. What then is the explanation for the discrepancy in these findings? The answer lies in the load (concentration x rate) of lactose that is administered to lactose-‘intolerant’ patients. Lactase, or fi-galactosidase, is a brush broder hydrolase whose specific activity is reduced but not absent in patients and subjects with lactose malabsorption. If 23 L lactose containing diet is infused constantly over 24 h, then the load of lactose administered per unit time is low, and symptoms, as demonstrated by Keohane and colleagues [lo] will not occur. Symptoms will develop, however, as demonstrated by O’Keefe and colleagues [27] if high loads of lactose are administered by the bolus technique. The commercial pressures are now such, that it is unlikely that any of the new enteral diets will contain appreciable quantities of lactose.

Sucrose

In patients with a very short small intestine, the factors. that will limit uptake of glucose from glucose polymer mixtures will usually include intraluminal and brush border saccharide hydrolysis and in a few instances when these are not rate limiting the capacity of the membrane carrier to mediate uptake of the released monosaccharide. Jejunal perfusion studies have shown that if glucose transport from glucose polymers is saturated, sugar absorption can be enhanced if the disaccharide sucrose is added [30]. This is so because sucrose is hydrolysed by the sucrase moiety of the hybrid brush border hydrolase sucrose-isomaltase [31] to glucose and fructose, fructose absorption then being mediated by a carrier that is distinct from that utilised during glucose uptake [32].

Protein absorption The digestion and absorption of dietary protein in relationship to the formulation of enteral diets has been recently reviewed [33]. While it remains clear that patients with normal gastrointestinal function should be fed diets containing whole protein as the nitrogen source, controversy still exists as to whether the nitrogen source of enteral diets for use in patients with severly impaired gastrointestinal function should consist of free amino acids or oligopeptides, and if the latter, what was the most rational formulation should be. Intestinal perfusion studies in our laboratory have consistently shown more efficient absorption of Xamino acid nitrogen from partial enzymatic hydrolysates of whole protein than equivalent equimolar free amino acid mixtures [34]. On the basis of these physiological studies we have proposed theoretical reasons for using peptides rather than free amino acids as the nitro-

64

TOWARDS THE OPTIMIZATION

OF ENTERAL NUTRITION

gen source in predigested chemically defined ‘elemental’ diets [4]. Other benefits which include effects on osmolality, cost and palatability have also been pointed out [4, 331. We have been struck by the variability in handling of the different partial enzymic hydrolysates of whole protein studied [34] and in recent studies have elucidated the influences that starter protein composition, hydrolysis method [35] and peptide chain length [36] have an absorption profile of partial enzymic hydrolysates of whole protein. One of the most interesting observations has been the finding that subtle increases in peptide chain length from 2-3 to 3-5 amino acid residues have a significantly deleterious effect on x-amino acid nitrogen absorption [37]. If a peptide based nitrogen source is considered desirable, evidence now points to the fact that mixtures of di- and tripeptides rather than higher peptides will result in the most efficacious rates of a-amino acid nitrogen absorption per unit length of intestine. Amino acid versus peptides-clinical

studies

There have appeared recently three studies [38-40] that have compared the efficacy of Vivonex HNa,, whose nitrogen source is composed of free amino acids, and Criticare HN” whose nitrogen source is composed of partially hydrolysed protein (peptides and amino acids). In the first Smith and colleagues [38] demonstrated higher blood urea nitrogen and urinary urea excretion values in patients with inflammatory bowel disease fed Vivonex HN K. In the second, similarly increased blood urea nitrogen values were observed in patients with pancreatic insufficiency fed with the two diets [39]. A significant weight gain was seen only in the patients fed Criticare HN” [39]. In malnourished patients with head and neck cancer significantly greater weight gains, higher serum albumins and lower rises in blood urea nitrogen were seen in the patients fed with Criticare HNH [40]. Contrary to some unpublished suggestions, the differences noted in these three trials do not indicate a superiority of peptides over amino acids. The quantitative differences of the amino acid composition of the two diets is marked [40], particularly in respect of glutamine an amino acid known to directly stimulate the urea cycle [40]. Ureagenesis with resultant increases in excretion of urinary urea nitrogen and impairment of nitrogen balance would thus be expected to occur with Vivonex HNx as compared to Criticare HNH. In our unit, we have also noted higher urinary urea excretion values in patients fed Vivonex’h, as compared with Clinifeed 40012, a whole protein containing polymeric diet [9]. The question as to whether peptides have a specific nutritional advantage over free amino acids can only be answered if patients with severely im-

paired gastro-intestinal function are fed peptide and free amino acid diets of identical amino acid composition. The preliminary results of such a study performed in our unit are now available. No significant differences in protein turnover rates were observed in patients with inadequate short bowel syndrome [41] fed a peptide as compared to a free amino acid containing diet. Unfortunately chromatographic analysis of peptide chain lengths in the former diet showed that significant quantities of peptides containing four or more amino acid residues were present. Hydrolysis of these by brush border peptidases limits the rate of amino acid nitrogen absorption [37] so that the study will have to be repeated using a defined di- and tri-peptide containing diet before final conclusions about the possible nutritional superiority of peptide containing predigested diets can be drawn.

Vitamin and trace elements The mineral, trace element and vitamin contents of enteral diets have traditionally been based on recommended dietary allowances (RDA) which are the levels judged by the various National Boards as well as WHO on the basis of available scientific knowledge to be adequate to meet nutritional needs of practically all healthy persons. As has been pointed out [42], the RDA do not cover therapeutic nutritional needs, so that vitamin and trace element requirements as affected by disease or pharmaceutical preparations may not be met in all enteral feeds [42]. As far as vitamin requirements are concerned, the subject has been comprehensively reviewed [42]. Of the trace elements, zinc has received the most attention and in 2-3 times the RDA is recommended for hypermetabolic patients [43]. Requirements of other trace elements for hypermetabolic patients are unknown, and deficiencies of magnesium, copper, chromium and selenium have all been reported in patients receiving parenteral nutrition [44-46]. Recently Bunker & Clayton [47] have reported discrepancies between levels of zinc, copper, iron and manganese in enteral diets stated by the manufacturers and those found on analysis. Several diets were actually found to contain less than the United States RDA [47].

Electrolytes Despite the fact that nearly 9 L of water and in excess of 1000 mmol sodium is ingested, secreted and absorbed, little emphasis has been placed, until recently, on water and electrolyte absorption during enteral feeding. Small intestinal disease or resection results in an increased colonic influx of water and electrolytes. As a

CLINICAL NUTRITION

consequence

of the large absorptive

capacity of the

colon [48], however, severe diarrhoea and depletion of water and electrolytes usually only occurs in the presence of co-existing colonic disease or resection. It follows that the ‘diarrhoea’ that occurs in enterally fed patients with normal or near normal gastrointestinal function in up to 2494 of cases [9], is either due to a hormonally induced colonic secretion of water and electrolytes, a hypothesis currently unsubstantiated [49], or to other causes such as a synergistic action between enteral feeding and concomitant antibiotic therapy [50]. In patients with the short bowel syndrome, however, the situation is quite different. Here one is faced with the situation where large segments of small bowel are diseased or have been resected, often with co-existing colonic disease or resection. In the absence of the reserve absorptive capacity of the colon [48], it becomes imperative that fluid and electrolyte absorption from residual small gut is maximised during enteral feeding. We were disturbed to find, therefore, that a net secretion of fluid and electrolyte was observed during in vivo steady state perfusion of segments of normal jejunum with isotonic Vivonex” [39]. Further studies have been performed using a range of solutions containing the amino acid and glucose polymer components of VivonexH [51]. Results showed that for all the nine nutrient solutions studies there was a linear correlation between initial sodium concentration and net sodium absorption so that if the initial concentration exceeded 90mmol/l net sodium absorption occurred, while at concentration below this, net secretion of sodium was observed. Water absorption was linearly related to net absorption of osmotically active particles, a substantial proportion of which was accounted for by the movement of sodium and its anions. Net sodium secretion observed during perfusion of Vivonex” [30] is not a unique feature of this diet, but will occur during jejunal perfusion of any low sodium containing diet. The sodium content of four widely used predigested chemically defined ‘elemental diets’, Vivonex b, Vivonex HN”, Nutranel* and Flexical” is 39.2, 36.3, 20.1 and 30.1 mmol/l respectively, values substantially lower than would be required to promote net sodium and water absorption. There are thus proven experimental grounds for suggesting that the sodium concentration of the predigested diets, when used in the management of the inadequate short bowel syndrome should be raised to 90 mmol/l.

Choice of enteral diets’ Patients function

with

In a carefully

normal or near normal gastrointestinal

performed

study,

Moriarty

and

col-

65

leagues [52] showed that no significant differences occurred in nitrogen balance when patients with normal gastrointestinal function were fed three enteral diets of similar composition excepting the nitrogen source which was composed either of whole protein, oligopeptides or free amino acids. There would thus seem to be no advantage in feeding patients with normal gastrointestinal function with a predigested chemically defined diet. That these patients should receive a polymeric diet is further supported by the results of two controlled clinical trials [9,53]. As mentioned above the formulation of these diets is based on our knowledge of the physiology of nutrient absorption and what is known about nutrient requirements in health and disease. The composition of two polymeric diets is summarised in Table 4. Until recently we have recommended the routine use of a polymeric diet of energy density 1 kcal/ml for non hypermatabolic patients [4]. It has been a clinical observation in our unit that it is difficult to actually administer more than 2-2.5 L of enteral diet per day to the routinely word fed patient [9, 10, 501. In the light of this we have recently tested a hypothesis that nitrogen balance in routinely ward fed patients with normal or near normal gastrointestinal function can be improved by administering an energy and nitrogen dense polymeric diet. In a double blind randomised controlled clinical trial [53] significantly better nitrogen balance was seen when an energy nitrogen dense diet (1.5 kcal/ml; 9.4g N/l) was administered as compared to either an energy dense (1.5 kcal/ml; 7.8g N/l) or a standard (1 kcal/ml; 6.3g H/l) polymeric diet. As a result of these findings we believe that a radical rethinking about polymeric diet formulation is required. We are now using energy and nitrogen dense diets for routine enteral feeding as they appear to be well tolerated [53], and lead to a greater efficacy of enteral nutrition at least as far as nitrogen balance is concerned. It can be seen from Table 4 that the osmolality of the energy and nitrogen dense diet is higher than the standard 1.0 kcal/ml diet. However, if the carbohydrate energy moiety was rationalised according to the concepts of purified high MW glucose polymer mixtures proposed by Jones and colleagues [24], the osmolality could be lowered. For very hypermetabolic patients, for example those with burns and multiple trauma with normal gastrointestinal function, polymeric diets are available with an energy density of 2.0 kcal/ml and nitrogen contents of up to 14g N/l. These are hypertonic diets and we have no personal experience of their uses. It should be borne in mind that there may be a real risk of infusing excess carbohydrate to those patients, some of whom already have insulin resistance. Furthermore, excessive carbohydrate loads could also have a deleterious effect

66

TOWARDS

Table 4

THE OPTIMIZATION

OF ENTERAL NUTRITION

Choice and formulation of enteral diets Polymeric diets for patients with normal or near normal gastrointestinal function Non-hypermetabolic hypermetabolic

to moderately

Current

Proposed

Hypermetabolic

Nitrogen source (g/l)

Protein

Protein

Protein

Carbohydrate

(5-7) Glucose polymers

P-10) Glucose polymers f sucrose

Fat source

Long chain triglycerides

Energy (9,) Kcal/ml Electrolytes (mmol/l) Sodium Potassium Chlorine Minerals (fraction BDA) Vitamins (fraction BDA) Osmolality (mosmol/kg)

(11-13)

Predigested diets for patients with severely impaired gastrointestinal function Purified low molecular weight peptide mixtures (5-10)

Purified glucose polymers (now > 10 glucose molecules) f sucrose

Glucose polymers

Long chain triglycerides

Long chain triglycerides

32-36

32-36

34-41

Medium chain triglycerides linoleic acid ?

1.0

1.5

2.0

1.0

30-70 30-70 30-70 1.0 1.0 300-400

30-70 30-70 30-70 l-l.5 l-l.5 up to 500

30-70 30-70 30-70 1.5-2.0 1.5-2.0 Up to 600

7c-90 30-70 70-90 1.0-l 5 1O-l.5 450-650

Table 4 is reproduced from Silk D B A. Diet formation and choice of enteral diet. Gut 1986,27: 4H6, publishers.

on respiratory function [54]. Modular formulae, whereby the composition of a diet is formulated in the diet kitchen according to the needs of the individual patient, has been proposed for these difficult patients 1551. Patients with impairedgastrointestinalfunction

In those clinical situations where the rate of nutrient absorption is limited by impaired iuminal hydrolysis, or reductions in mucosal absorption or hydrolytic capacity, the use of enteral diets containing predigested nutrients will result in more efficient nutrient repletion that the polymeric diets discussed above. Strictly speaking, these criteria for using the predigested or so called chemically defined elemental diets, will usually occur only in patients with severe exocrine pancreatic insufficiency and patients with an inadequate or short bowel syndrome. Although other indications for using predigested chemically defined elemental diets have been proposed [56], there is very little controlled evidence to support the majority of the claims [57]. When nutrient assimilation is severely impaired, there are theoretical reasons for believing that nutrients should be presented to the mucosa in the form that re-

by kind permission of the

sults in the maximal absorption in normal subjects. The suggested formulation of such a diet is summarised in Table 4 and is based on discussions of the physiology of nutrient absorption outlined at the beginning of this paper. In practical terms, it may often be difficult to decide whether a polymeric or predigested diet should be used. In the author’s opinion, the assimilatory capacity of the human gastrointestinal tract for nutrients is often under estimated. For example, there is no evidence that there is any clinically significant impairment of exocrine pancreatic secretion or intestinal absorptive capacity in the post-operative period. Moreover, there is no clinical evidence to support a contention that the use of predigested diets in jejunostomy feeding results in more efficient nutrient repletion than polymeric diets [58]. Little attention has been directed towards enteral feeding in patients with cholestatic jaundice. In the absence of other co-existing gastrointestinal disease these patients will be able to assimilate all the major classes of nutrients except fat and fat soluble vitamins. Diets for those patients should not contain LCTs, so most of the currently available diets are not indicated. A modular formula based on whole protein, glucose polymers, MCTs, linoleic acid (in the hope of preventing essential

CLINICAL NUTRITION

fatty acid deficiency), electrolytes, minerals and water and fat soluble vitamins would be most appropriate. MCT assimilation by the jaundiced patient will depend on a number of factors, including the degree of jaundice, and the presence of co-existing gastrointestinal pathology. Although there is no data to support it, MCTs should probably not contribute to more than 20”, of the total energy content of these formulations.

Chronic portal systemic encephalopathy Patients with chronic portal-systemic encephalopathy (PSE) are faced with a nutritional dilemma. They require dietary protein to maintain the nitrogen balance, but the ingestion of protein often precipitates encephalopathy. Despite treatment with lactulose, lactitol or neomycin, many cirrhotic patients are unable to tolerate sufficient protein to prevent long standing negative nitrogen balance. As the plasma amino acid profiles of these patients are characterised by decreased levels of branched chain amino acids (BCAA’s) and increased levels of aromatic amino acids (AAA’s) {phenylalanine, tyrosine, free tryptophan and methionine) [59] investigators have administered mixtures of amino acids rich in BCAA’s and deficient in AAA’s [60]. Overall, it has not been easy to draw conclusions from the published data since the controlled trials evaluated different types and degrees of encephalopathy using different mixtures of BCAA’s administered in different ways, using different end points and finding different results. As far as enteral nutrition is concerned, one trial [61] clearly shows that a BCAA enriched enteral diet low in AAA’s induced positive nitrogen balance in protein intolerant cirrhotic patients all of whom had had chronic PSE, to approximately the same degree as an equivalent amount of dietary protein without inducing encephalopathy as frequently. This interesting topic is discussed in more detail elsewhere [62,63].

Respiratory and cardiac failure In an important study, respiratory, cardiovascular and metabolic changes were monitored during balance studies in undernourished patients receiving continuous feeding [54]. At increasing energy infusion rates 0, consumption, CO, production, minute ventilation, heat production, heat release and heart rate increased on both a high carbohydrate and high fat diet. Rather disturbingly increases in CO, _-production and minute ventilation were greater for the high carbohydrate relative to the high fat formula. In the clinical setting, care must be taken not to precipitate heart failure during continuous enteral feeding in compromised patients. Further-

67

more, these findings also suggest that the ratio of carbohydrate to fat should be decreased in enteral diets administered to patients with respiratory failure. A new diet (Pulmocare,” Ross) has been formulated on these lines.

Stress There has been much interest in the role of BCAA’s in the management of the stressed patient. The plasma concentrations and catabolism of the BCAA’s change under stress [60] and as in vitro incubations of isolated muscle indicate that leucine specifically increases the rate of protein synthesis and decreases the rate of protein degradation [64, 681, a special role for BCAA’s as anticatabolic amino acids in stressed patients has been proposed. It can be argued, however, that the in vitro studies were grossly unphysiological as all preparations were in negative nitrogen balance (degradation > synthesis). In vivo leucine has no effect on muscle protein synthesis, even at greatly increased concentrations [69] and it is thus hard to justify the use of BCAA enriched formulations in enteral nutrition. Although Cerra and colleagues [70] showed that enriched parenteral solutions seemed to give a more rapid return to a positive nitrogen balance than standard solutions after abdominal surgery, the cumulative nitrogen balances measured over 6 days were the same. In their most recent study [71], BCAA enriched enteral diets resulted in better nitrogen balance after feeding for 5 and 7 days than standard diets in moderately high ICU stresss ( < 10 g urinary nitrogen/day). Any potential clinical benefits were not reported. The remaining controlled trials published recently were disappointing. Of the seven studies, one has shown greater nitrogen loss with high BCAA’s [72] and six have shown no difference in protein metabolism and nitrogen balance [73-781. The current vogue for using BCAA enriched formulations for stressed or catabolic patients is therefore probably misconceived for two reasons. Firstly, the metabolic rationale for high concentrations of BCAA’s and a specific role for leucine are questionable, Secondly, there has been no convincing proof over the past few years of a definite clinical benefit from these solutions.

Techniques

ofadministration

The practical aspects of administration techniques used during exiteral feeding are important. In reality, we believe we have increased the efficacy of enteral nutrition to a greater extent by modifying and improving our

68

TOWARDS THE OPTIMIZATION

OF ENTERAL NUTRITION

administration techniques than by altering diet formulation. Put very simply, there seems little point in advocating a change in diet formulation which results in a 5loo,; increase in nutrient assimilation if the administration technique being used results in the administration of only 60° u of prescribed diet [lo].

Feeding tube design and performance As Bastow has pointed out [79] when wide bore tubes of the Ryle type are used for nasogastric feeding, the size and rigidity of the tubes tend to produce irritation and inflammation of the oesophagus with subsequent haemorrhage and stricture formation. Such side effects are not seen with the newer softer and narrower bore tubes now available. We developed a special interest in enteral feeding tube design and performance when our first retrospective survey [80] highlighted the limitations and drawbacks of the simple open ended unweighted polyvinyl chloride nasogastric feeding tubes that we had used in 200 patients. On average 2.5 tubes were used per course of enteral nutrition and the mean duration of tube use was only 5.8 days. The non elective removal rate was high at 61.876. Although claims had been made that 3.5 g weights incorporated into the tip of nasogastric tubes act to keep the tube in the stomach and thus inhibit inadvertent displacement [81, 821, we could not confirm this is one retrospective and two subsequent controlled trials [80, 83, 841. The reasons for the high inadvertent removal rates have not been defined and to date we have not confimed assertions [85] that inappropriate removal is confined to forceful removal of tubes soon after placement by mentally confused patients [83]. Disappointed with the overall performance of the weighted and unweighted nasogastric tubes investigated in the first two of our studies [80, 831, a design programme was initiated which has resulted in the development of a new generation of polyurethane feeding tubes [84]. Polyurethane was preferred to PVC, silicone or latex because our clinical experience showed that ‘kink resistance’ could be maintained despite using thinner tubing. Thus while the outside diameter of the new tubes is the same as that of the PVC tubes, the internal diameter is 24.5% larger with a 54qb increase in flow area, providing, therefore, less resistance to infusion of the relatively more viscous energy dense feeds. The interior wall, as well as the outside of the tip of the feeding tube, is impregnated with a water activated lubricant, which eases tube insertion through the nasopharynx and facilitates the removal of the introducer wire. In the early phase of our tube development programme, difficulties arose when attempting to aspiration from side ports, particularly when multiple ports

were employed. Moreover, difficulties with diet infusion also arose when mucus and/or curdled diet accumulated in the tube distal to the side port or ports. In retrospect we realised that resistance to outflow could occur when too small a side port was employed, and that the shape was important, outflow seeming to be restricted if the edges of the port were rightangled. In the final design of the outflow ports of the new tubes, therefore, attempts were made to mimic as far as possible the flow characteristics of the open ended PVC tubes. Thus the tubes contain, a long, single, wide-necked, smooth and curved edged outflow port and there is no ‘dead space’ distal to outflow port to trap mucus or curdled diet. The performance of the new tubes has been investigated under prospective controlled trial condition and has been found to be superior to that of the previously used PVC tubes used as a control [84], in regard to duration of tube use, ease of intubation and in the ability to aspirate gastric contents after tube insertion. The delay that occurs before reintubating patients once inadvertent tube removal has occurred [83] leads to inevitable reductions in nutrient intake which in turn impairs the efficacy of enteral nutrition. Thus while we consider that the new polyurethane nasogastric feeding tubes (Corpak Co, Wheeling, IL, USA) represent a significant advance, further research is clearly needed to develop a masogastric tube that maintains its position for longer in the stomach of both conscious and unconscious enterally fed patients.

Diet containers Concern has been expressed over the possible role of nasogastric feeds as potential sources of infection and cross infection, especially in patients in intensive care units [86], and the whole subject of microbial contamination of enteral feeds has now been reconsidered [87]. The concern about diet contamination led to the widespread use of 500 ml diet containers that required changing at least every 6 h. Bastow and colleagues [88] have shown that if sterile commercial feeds are carefully emptied into diet containers on the ward they remain sterile, whereas if a diet is blended with additives in the diet kitchen it is likely to become contaminated and subsequent bacterial multiplication occurs ( 109*0 organisms/ml after 24 h exposure to ward temperature (21-24°C)). As a consequence of this, it is clear that if enteral feeds are to be prepared in the diet kitchen then 500ml diet containers must be used. If pre-sterilised enteral diets are used without blending, then 1.5-2 L containers can be used. We have recently shown in two separate controlled clinical trials [ 10, 891 an added and important advantage of using large volume diet con-

CLINICAL NUTRITION

tainers. Thus significant greater proportions of the daily prescribed diet were actually administered to the patient from a single 2 L container than from four 0.5 L or two l.OL containers [lo, 891. The single 2 L containers used were pre-packaged with presterilised polymeric diet and were preferred by the nursing and dietetic staff as they took less time to prepare [893. A word of caution is required, however, as one of 648 2 L containers used in the second study [89] was found to be contaminated before use, and two of 26 containers from which aliquots were cultured during use, showed significant (> lo* organism/ml) bacterial contamination. More stringent nursing procedures are probably required during the setting up of the infusion system, and possibly further thought needs to be given to the type of material used to manufacture the 2L prepackaged diet containers before this type of system finds widespread clinical usage. As far as cost effectiveness is concerned, it must be remembered that it is the water component of enteral diets that is the most expensive to transport, and the most bulky component to store. As mentioned previously [90] we believe that the most efficient system in the future may be the distribution of 2 L bags containing pre-sterilised powder which is then reconstituted with water just prior to use. Such a system is currently being evaluated in our unit.

Routes of administration Nasogastric

zj nasoenteral

tube feeding

The main disadvantage of nasoduodenal or nasojejunal feeding is that the pylorus is bypassed. Gastric emptying is mediated by the action of the pylorus, and the mechanisms underlying the ‘duodenal braking effect’, whereby the rate of gastric emptying is governed by entry of gastric contents into the duodenum, have been analysed recently by Spiller [91]. The importance of the ‘duodenal brake’ is often ignored in enteral nutrition, and the point that needs emphasising for the future is that the osmotic load of nutrients presented to the duodenum does not depend on the product of diet osmolality and rate of diet infusion, but on the product of gastric effluent osmolality and rate of gastric emptying. The cramping, distension, and diarrhoea that occur during nasoenteric feeding is probably related, at least in part to the rapid secretion of fluid and electrolytes in response to the high osmotic load of nutrients entering the upper small bowel. The ‘duodenal brake’ will reduce these symptoms if enteral feeds are infused intragastrically. In our experience, using the nasogastric rather than the nasoenteral route reduces the necessity of slowing infusion rate to counteract the development of gastrointestinal side effects. In turn this has the effect

69

of optimising nutrient intake and thus efficacy of enteral feeding. Although there have been no controlled clinical trials to support the theory, it does seem likely that regurgitation and aspiration of feed occur more commonly in nasogastrically than nasoenterically fed neurological patients. Particularly at risk, it is said, are the aged, debilitated, demented and stuperous patients; those with poor gag reflexes and neuromotor deglutition disorders; and neurosurgical patients in the immediate postoperative phase [92]. A case then, therefore, can be made to feed this subgroup of patients nasoenterally. Despite the fact that at least 30”+, of our enterally fed patients have underlying neurological or neurosurgical disease, our documented incidence of aspiration (nine cases out of 781 patients enterally fed over 8536 days [83]) is lower than that of others [93]. Since we believe that greater nutritional intakes can be achieved via the nasogastric than nasoenteral route we usually resort to nasoenteral feeding in only the very difficult cases. Routes for long term enteral feeding It is possible that as physicians my group has a propensity for advocating long term enteral feeding via the nasogastric route rather than via a surgically created gastrostomy or jejunostomy. Bearing in mind the limitations of nasograstric feeding tubes discussed above it seems possible that the efficacy of long term enteral feeding could on occasions be improved by resorting to the surgical creation of a more permanent route of administration. The techniquess of surgically creating tube enterostomy feeding routes have been eloquently summarised by Rombeau and colleagues [94]. With respect to gastrostomy feeding, a new advance in the surgical technique has been described by Sriram and colleagues [95]. Based on the construction of a permanent mucosal-lined gastrostomy using a stapling device, the gastronomy is essentially ‘continent’, that is the feeding tube can be removed in between feeding with no leak of gastric contents. This in turn leads to a lower incidence of skin excoriation, a well recorded complication of traditional gastrostomies [96]. Tubes used for enterostomy feeding are well known to have their limitations. Ongoing research is in progress to redesign these tubes in the hopes of improving performance. As with the nasogastric tubes, the current prototypes are manufactured with polyurethane and have similarly designed exit ports (D. Quinn, unpublished observations). As an alternative to routes of administration created by formal surgical techniques, a number of percutaneous endoscopic techniques of fashioning a tube feeding gastrostomy have recently been described, and

70

TOWARDS

THE

OPTIMIZATION

OF ENTERAL

NUTRITION

recent experiences with these techniques have been reviewed [97]. Although increasing experience is being gained with this type of technique in America, we have only a very limited and rather unsuccessful experience ourselves. Nevertheless such approaches are likely to gain a much wider application in the future, particularly when enterostomy feeding tube design has been improved. In difficult cases where aspiration of enteral diet is almost inevitable and in patients requiring long term enteral feeding, nutritional intake is likely to be enhanced using this technique, in turn leading to improved efficacy of enteral feeding.

Starter regimes Gastrointestinal side effects that occur during enteral nutrition include nausea, abdominal bloating and pain as well as diarrhoea. Diarrhoea, defined as the passage of too loose and frequent stools to be of discomfort to the patient or the nursing staff, occurs in up to a quarter of patients receiving enteral nutrition [9]. Traditionally, it has been believed that the incidence of these side effects, particularly diarrhoea can be minimised by gradually introducing full strength feeding over a 34day period at the outset of feeding-the so called ‘starter regimen’ technique. On the basis of this thinking one presumed that the onset of side effects, particularly the diarrhoea, was related to initial intolerance of the high osmotic loads presented to the upper small intestine during enteral feeding. Unimpressed by such thinking Keohane and colleagues, undertook a prospective randomised controlled clinical trial to examine the relationship between the onset of gastrointestinal side effects and diet osmolality [50]. Patients with normal gastrointestinal function requiring constant 24 h nasogastric infusion of a polymeric diet were randomised to receive either undiluted, ‘full strength diet’ from the outset, the same diet but using a ‘starter regimen’ whereby 4 days were taken to gradually introduce full strength feeding, or an isotonic diet. The incidence of gastrointestinal side effects was somewhat lower rather than higher in the first group of patients receiving full strength diet from the outset. While nitrogen losses were similar in all three groups, nitrogen intake was significantly greater in the patients receiving full strength diet from the outset, the net result of this being significantly better nitrogen balance in this group of patients. It was entirely clear from these data that the efficacy of enteral feeding could be improved by not using starter regimens and since completion of the above trial [50] the routine use of starter regimens in our unit has been abandoned. Further studies have shown that the relatively hyperosmolar predigested ‘chemically defined elemental’ diets can also be routinely administered to

patients with acute bowel disease without

exacerbations of inflammatory ‘starter regimes’ [98].

Complications of enteral nutrition In our experience the efficacy of enteral nutrition can be maintained even in the event of many of the known side effects developing. This is so because once an understanding is gained of the factors involved in their pathogenesis, the necessary steps can be taken to deal with the problem, often without reducing the rate of diet infusion. It is necessary to stop enteral feeding on account of the onset of side effects only very rarely. The common complications of enteral nutrition are summarised in Table 5. Probably the commonest is diarrhoea, occurring as mentioned above in up to a quarter of patients [9]. A number of factors have been implicated in its pathogenesis [4]. These include the use of infected feeds, lactose intolerance, intolerance of high osmotic loads administered, inappropriate release of gastrointestinal polypeptide hormones, concomitant antibiotic therapy, and ingestion of laxatives. If an unexpected outbreak of diarrhoea occurs, infection of water used for diet dilution or reconstitution or powder should be suspected and water cultured [9]. The relevance of lactose intolerance has been discussed above, as has the question of diet osmolality. Animal studies performed in our laboratory have not confirmed that the onset of diarrhoea is related to an inappropriate release of gastrointestinal polypeptide hormones [49]. We have identified laxatives as a cause of diarrhoea in some patients [9]. In the majority of patients, diarrhoea occurs in association with concomitant antibiotic therapy administered either by the oral or parenteral route [50]. It seems that there is a curious synergistic effect of enteral feeding and antibiotic therapy. Intestinal perfusion studies performed in our unit suggest that certain antibiotics have a deleterious effect on fluid and electrolytes assimilation in the human small intestine [99], which might have relevance to this problem. We have speculated that in at least some patients the diarrhoea may have a colonic basis. Short chain fatty acids (SCFAs), normally produced in the colon as a consequence of the bacterial metabolism of unabsorbed carbohydrate and fibre, act as a powerful stimulant to colonic water and electrolyte absorption [ 100, lOl]. In our unit we have developed an ‘artificial colon system’ whereby we are able to assess the effect of normal human faecal bacterial metabolism in vitro [ 1021. Certain antibioticss inhibit SCFA production (RC Spiller & DBA Silk, unpublished observations). Enteral diets have a very low fibre content and investigations are currently in progress in our laboratory to assess the completeness of small intestinal carbohydrate assimilation during enteral feeding. We are therefore currently test-

CLINICAL NUTRITION

Table 5 Complications

71

of enteral nutrition

Type

Comment

Remedy

Feeding tube related Oesophageal inflammation haemorrhage or stricture formation

Rare with new ‘fine bore’ PVC or polyurethane tubes

Prophylactic treatment with H, receptor antagonists if wide bore Ryle type tubes have to be used for more than 4-5 days

Tube misplacement

Most commonly patients

Check position by X-ray

Oesophageal, nasopharyngeal or gastric perforation

Rare

Can be treated conservatively

Occurs transiently in 5-15% of patients; not usually severe. Probably due to gastric stasis and high osmotic loads of nutrients administered

Usually settle spontaneously. Occasional need to reduce rate of diet infusion

Diarrhoea

See text

If possible stop antibiotics treat with imodium or codeine phosphate

Biochemical hyperglycaemia

Usually

Nausea Abdominal Abdominal

bloating pain

I

occurs

in unconscious

Related to nitrogen anabolic phase

losses or occurs

hypophosphataemia

Related

losses

vitamin, mineral, trace elements, essential fatty acid deficiencies

Rare. Monitor

Abnormal

liver function

ing a hypothesis feeding related creased colonic combination of fibre substrates by antibiotics.

tests

Insulin

related to insulin resistance

hypokalaemia

to nitrogen

clinically

in

Supplements required Supplements required

and biochemically

Causes multifactorial and related to underlying disease or malnutrition. May in part be due to continuous 24 h infusion of diet

that there is a colonic basis to enteral diarrhoea which in turn is related to deSCFA production, consequent upon a a reduction in colonic carbohydrate and and inhibition of bacterial fermentation

Conclusions This article has attempted to summarise the progress we had made towards optimising enteral nutrition. In the light of our experiences we would recommend that patients with normal or near normal gastrointestinal function should receive 2-2.5 L correctly formulated polymeric diet (1.5 kcal/ml, 9 g N/L) via continuous 24 h infusion (without a starter regimen) from 1-2L containers using unweighted fine bore polyurethane feeding tubes incorporating an anti-clog tip situated wherever possible in the stomach. The same techniques should be used to infuse correctly formulated predigested chemically defined elemental diets to patients with severely impaired gastrointestinal function. Patients should be carefully monitored, clinically, biochemically and haematologically in order to identify

Supplements required

In our experience seldom clinically significant. In patients without liver disease, LFTS return to normal after cessation of feeding

side effects early, most of which can be dealt with early, without prejudicing the efficacy of enteral feeding. ACKNOWLEDGEMENTS None of the work described

in this paper would have been possible without the scholarly efforts and hard work on the part of my past and present research fellows. My grateful thanks are due to Dr B. J. M. Jones, Dr P. P. Keohane, Dr R. C. Spiller, Dr D. H. Patil and Dr R. G. Rees. The analytic laboratory in my unit is run by Dr G. K. Grimble. His innovative work and efforts are also gratefully acknowledged. Miss Helen Attrill, Nutrition Sister, has played a major role in the organisation and performance of the clinical trials. Her efforts have been much appreciated. My secretary, Miss Tina Carrick, has made a major contribution throughout and we are all very grateful to her. Finally, the financial contributions received from the nutrition industry have enabled our work to proceed. All those who have contributed have done so in a most helpful way, providing at all times stimulatory and helpful ideas, which have permitted a fruitful and ongoing collaboration between industry and our unit.

REFERENCES

[l] Bristian B R, Blackburn

G L, Hallowell F, Heddle R 1976 Prevalence of malnutrition in general medical patients. VAMA 235: 1567-1570

72

TOWARDS THE OPTIMIZATION

OF ENTERAL NUTRITION

[2] Bristian B R, Blackburn G L, Hallowell E, Heddle R 1974 Protein status of general surgical patients. Journal of the American Medical Association 230: 858-860 [3] Sutton G, Karran S T 1985 The diagnosis of malnutrition: Nutritional parameters. Current Opinion in Gastroenterology 1: 281-287 Gower Academic Medical Journals, London [4] Silk D B A 1983 Nutritional Support in Hospital Practice. Blackwell Scientific Publications, Oxford [5] Phillips G D, Odgers C L 1986 Parenteral and Enteral Nutrition. A Practical Guide, Churchill Livingstone, London (61 Carpentier Y A 1986 Indications for nutritional support. Gut 27: Supplement I, 1417 [7] Jones B J M, Payne S, Lees R et al 1980 Indications for pump assisted enteral feeding. Lancet 1: 1057-1058 [S] Jeejeebhoy K N 1986 Muscle function and nutrition. Gut 27: Supplement I, 25-39 [9] Jones B J M, Lees R’, Andrews J et al 1983 Comparison of an elemental and polymeric enteral diet in patients with normal gastrointestinal function. Gut 24: 78-84 [lo] Keohane P P, Attrill H, Love M, Frost P, Silk D B A 1983 controlled trial of aseptic enteral diet preparation-significant effects on bacterial contamination and nitrogen balance. Clinical Nutrition 2: 119-122 [ 1 l] Jones B J M, Silk D B A 1982 Parenteral nutrition. Medicine International 15: 674-678 [12] Casewell M W 1979 Nasogastric feeds as a source of Klebsiella infection for intensive care patients. Research & Clinical Forums 1: 101-105 [ 131 Keighley M R B, Mogg B, Beatley S, Allan C 1982 ‘Home brew’ compared with commercial preparations for enteral feeding. British Medical Journal 1: 163 [ 141 Dodge J A, Yassa J G 1980 Essential fatty acid deficiency after prolonged treatment with elemental diet. Lancet ii: 1256-1257 [15] Silk D B A 1984 Future directions in supplemental nutrition. In: Cystic Fibrosis. Horizons. D Lawson (ed.) John Wiley & Sons, Chichester pp 99-l 14 [ 161 Hashim S A 1967 Medium chain triglycerides: clinical and metabolic aspects. Journal of the American Dietetic Association 57: 221-227 [17] Forstner G, Gall G, Corey M et al 1980 In: Proceedings of the 8th International Congress of Cystic Fibrosis, Toronto. Canadian Cystic Fibrosis Foundation pp 137148 [ 181 Moran J R, Greene H L 1984 Digestions and absorption in Enteral and Tube Feeding 1984. J L Rombeau M D Caldwell (eds) WB Saunderss Co, Philadelphia pp 20-43 [ 191 Greenstein J P, Otey M C, Birnbaum S M et al 1960 Quantitative nutritional studies with water soluble, chemically defined diets. X Formulation of a nutritionally complete diet. Journal of the National Cancer Institute 24: 211-219 [ZO] Winitz M, Seedman D A, Graff J 1970 Studies in metabolic nutrition employing chemically defined diets. I. Extended feeding of normal human adult males. American Journal of Clinical Nutrition 23: 525-545 [21] Randall H T 1984 Enteral Nutrition. Tube feeding in acute and chronic illness. Journal of Parenteral and Enteral Nutrition 8: 113-l-136 [22] MacBumey MM, Young L S 1984 Formulas. In: Enteral and tube feeding. J L Rombeau, M D Caldwell (eds) WB Saunders Co, Philadelphia pp 171-198 (231 Jones B J M, Brown B E, Loran J S, Edgerton D,

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

Kennedy J F, Stead J A, Silk D B A 1984 Glucose absorption from starch hydrolysates in the human jejunum. Gut 24: 1152-l 160 Jones B J M, Brown B E, Spiller R C, Silk D B A 1981 Energy dense enteral feeds-The use of high molecular weight glucose polymers. Journal of Parenteral and Enteral Nutrition 5: 567 Neale G 1968 The diagnosis, incidence and significance of disaccharidase deficiency in adults. Proceedings of the Royal Society of Medicine 61: 1099-l 117 Walike B C, Walike J W 1977 Relative lactose intolerance: A clinical survey of tube-fed patients. Journal of the American Medical Association 238: 948951 O’Keefe S J D, Adam J K, Cakata E, Epstein S 1984 Nutritional support of malnourished lactose intolerance African patients. Gut 25: 942-947 Haverber L, Kwon P H, Scrimshaw N S 1980 Comparative tolerance of adolescents of differing ethnic backgrounds to lactose containing and lactose free dairy drinks. I. Initial experience with a double-blind procedure. American Journal of Clinical Nutrition 33: 17-21 Kwon P H, Rorick M H, Scrimshaw N S 1980 Comparative tolerance of adolescents of differing ethnic backgrounds to lactose containing and lactose free dairy drinks. II. Improvement of a double-blind test. American Journal of Clinical Nutrition 33: 22-26 Jones B J M, Brown B E, Silk D B A 1980 Isotonic vivonex causes water secretion in normal human jejunum. Gut 21: A920 Conklin K A, Yamashiro K M, Gray G M 1975 Human intestinal sucrase-isomahase: identification of free sucrose and isomaltase and cleavage of the hybrid into active distinct subunits. Journal of Biological Chemistry 270: 5735-5441 Milla P J, Oyesiku J E, Muller D P R, Harries J P 1977 Fructose absorption and the effects of other monosaccharides on its absorption in the rat jejunum in vivo. Gut 18: 425 Keohane P P, Silk D B A 1984 Peptides and free amino acids. In: Enteral and tube feeding. J L Rombeau, MD Caldwell (eds) WB Saunders Co, Philadelphia pp 44-59 Silk D B A, Fairclough P D, Clark M L et al 1980 Uses of a peptide rather than free amino acid nitrogen source in chemically defined elemental diets. Journal of Parenteral and Enteral Nutrition 4: 548-553 Keohane P P, Brown B E, Grimble G K, Silk D B A 1981 The peptide nitrogen source of elemental diets comparisons of absorptive properties of five partial enzymic hydrolysates of whole protein. Journai of Parenteral and Enteral Nutrition 5: 568 Grimble G K, Keohane P P, Higgins B E, Kaminski M V, Silk D B A 1986 Influence of peptide chain length on amino acid and nitrogen from two lactalbumin hydrolysates in the normal human jejunum. Clinical Science 7 1: 65-69 Rees R G, Grimble G K, Keohane P P, Higgins B E, West M, Spiller R C, Silk D B A 1984 Peptide chain length of protein hydrolysates influences jejunal nitrogen absorption. Gut 25: A 547 Smith J L, Arteaga C, Heymsfield S B 1982 Increased ureagenesis and impaired nitrogen use during infusion of a synthetic amino acid formula, a controlled trial. New England Journal of Medicine 306: 1013-1018 Nasrallah S M, Martin D M 1984 Comparative effects of Criticare HN and Vivonex HN in the treatment of

CLINICAL NUTRITION

[40]

[41]

[42]

[43]

[44]

(451

[46]

[47] [48]

[49]

[50]

[5 l]

[52]

[531

1541

[551

[561

malnutrition due to pancreatic insufficiency. American Journal of Clinical Nutrition 39: 251-254 Megvid M M, Landel A M, Terz J J, Akrabawi S S 1984 Effect of elemental diet on albumin and urea synthesis: comparison with partially hydrolysed protein diet. Journal of Surgical Research 37: 16-24 Grimble G K, Rees R G, Halliday D, Ford G C, Silk D B A Are enrerally fed peptides better utilised than free amino acids in the short bowel syndrome. Clinical Nutrition (abstract) (in press) Caldwell MD, Caldwell C K 1984 Micronutrients and enteral nutrition. In: Enteral and tube feeding. J L Rombeau, MD Caldwell (eds) WB Saunders Co, Philadelphia pp 84-126 Molnar S A, Bell S J, Goodenough R D, Burke J F 1984 Enteral nutrition in patients with bums or trauma. In: Enteral and tube feeding. J L Rombeau, M D Caldwell (eds) WB Saunders Co, Philadelphia pp 412433 Kay R G, Tasman-Jones C D, Pybus J et al 1976 A syndrome of acute zinc deficiency during total parenteral alimentation in man. Annals of Surgery 183: 331-340 A M Van Ri J, Thomson C D, McKenzie J M, Robinson M F 1979 Selenium deficiency in total parenteral nutrition. American Journal of Clinical Nutrition 32: 2076-2085 Freeman J B, Whittine M F 1977 Magnesium requirements are increased during total parenteral nutrition. Surgical Forum 28: 61-62 Bunker V W, Clayton B E 1983 Trace element content of commercial enteral feeds. Lancet 2: 426-428 Phillips S F, Giller J 1973 The contribution of the colon to electrolyte and water conservation in man. Journal of Laboratory and Clinical Medicine 8 1: 733746 Brown B E, Silk D B A 1982 Influences of duodenal Vivonex and Ensure infusion on ileal and colonic water electrolyte absorption. Journal of Parenteral and Enteral Nutrition 6: 321 Keohane I’ P, Attrill H, Love M, Frost I’, Silk D B A 1984 Relation between osmolality of diet and gastrointestinal side effects in enteral nutrition. British Medical Journal 288: 678-681 Spiller R C, Jones B J M, Silk D B A Jejunal water and electrolyte absorbtion from two proprietary enteral feeds in man-importance of sodium content. Gut (in press) Moriarty K J, Hegarty J E, Fairclough P D, Kelly M J, Clark M L, Dawson A M 1985 Relative nutritional value of whole protein, hydrolysed protein and free amino acids in man. Gut 26: 694-699 Rees R G I’, Cooper T M, Frost P G, Silk D B A 1986 Influence of energy and nitrogen balance: a double blind prospective controlled trial. Gut 27: A 1274 Heymersfield SB, Head A, McManus C B, Seitz S, Stations G W, Grossman G D 1984 Respiratory, cardiovascular and metabolic effects of enteral hyperalimentation: influence of formula dose and composition. American Journal of Clinical Nutrition 40: 116-130 MacBumey M M, Jacobss D 0, Apelgren K N et al 1984 Modular feeding. In: Enteral and tube feeding. J L Rombeau, M D Caldwell (eds) WB Saunders Co, Philadelphia pp 199-2 11 Russell R I 1975 Elemental diets. Gut 16: 68-79

73

[57] Koretzy R L, Meyer J H 1980 Elemental diets-Facts and fantasies, Gastroenterology 78: 393-410 [58] Fick Th E, Van Rooyen W, Eeftinck Schattenkerk M, Steller E Ph, Feenstra B W A, Bruining H A I986 A non elemental versus an elemental diet for early postoperative enteral feeding via needle catheter jejunostomy. Clinical Nutrition 5: 105-108 [59] Morgan MY, Milson J P, Sherlock S 1978 Plasma ratio of valine, leucine and isoleucine to phenylalanine and tyrosine in liver disease. Gut 19: 1068-1073 [60] Walser M 1984 Therapeutic aspects of branched-chain amino and keto acids. Clinical Science 66: 1-15 [61] Horse D, Grace N D, Corm H D, Schiff E, Schenker S, Viteri A, Law D, Atterbury C E 1984 Comparison of dietary protein with an oral branched chain-enriched amino acid supplement in chronic portal-systemic encephalopathy: A randomised controlled trial. Hepatology 4: 279-237 [62] Silk D B A 1986 Branched chain acid therapy in patients with cirrhosis and hepatic encephalopathy. Journal of Clinical Nutrition and Gastroenterology 1: 103-108 [63] Freund H R, Fischer J E 1986 The user of branched chain amino acids (BCAA) in acute hepatic encephalopathy. Clinical Nutrition 5: 135-138 [64] Buse M G, Jursinic S, Reid S S 1975 Regulation of branched chain amino acid oxidation in isolated muscle nerves and aortas of rats. Biochemical Journal 148: 363374 [65] Buse M G, Reid S S 1975 Leucine, a possible regulator of protein turnover in muscle. Journal of Clinical Investigation 56: 1250-1261 [66] Fulks R M, Li J B, Goldberg AL 1975 Effects of insulin, glucose and amino acids on protein turnover in rat diaphragm. Journal of Biol Chem 250: 290-298 [67] Li 1 B, Jefferson L S 1978 Influence of amino acid availabiiity on protein turnover in perfused skeletal muscle. Acta Biochemica et Biophysics 544: 351-359 K, Edstrom S, Ekman L et al 1981 Protein [681 Lundholm degradation in human skeletal muscle tissue: the effect of insulin, leucine, amino acids and ions. Clinical Science 60: 319-326 M A, Fern E B, Garlick I’ J 1982 Failure of 1691 McNurlan leucine to stimulate protein synthesis in vivo. Biochemical Journal 204: 831-838 [701 Cerra F B, Upson D, Angelic0 R et al 1982 Branched chains support postoperative protein synthesis. Surgery 92: 192-198 N N et al 1985 [7Il Cerra F B, Shronts E P, Konstantinides Enteral modified amino acids (MAA) are efficacious in ICU stresss. Journal of Parenteral and Enteral Nutrition 9: 124 [721 Lenssen P, Cheney C, Aker S et al 1985 High branched chain amino acid trial in marrow transplantation for leukaemia. Journal of Parenteral and Enteral Nutrition 9: 124 chain [731 Kirvela 0, Takala J 1985 The effect of branched amino acids and the keto-analogues on postoperative N balance and liver protein synthesis. Clinical Nutrition 4: (SupplO70) M F, Routlart M et al 1741 De Jong K P, von Meyendfeldt 1985 The effect of branched chain amino acids (BCAA) on organ functioning in septic and stressed patients. Clinical Nutrition 4: (SupplO71) J P, Wybran J 1985 [751 Gelin M, van de Stadt J, Lamillotte Humoral and cellular immunological response to

74

[761

1771

[781

I791 [801

[811

[821

J831

[841

f851

(861

[871

JSS]

TOWARDS THE OPTIMIZATION

OF ENTERAL NUTRITION

standard and BCAA enriched solutions of TPN. Clinical Nutrition 4: (SupplO72) Nuutinen L, Wistbaka J-O, Kaukinen S et al 1985 The comparison of ‘normal’, BCAA-enriched and intracellular amino acid change adapted amino solutions used after major abdominal surgery. Clinical Nutrition 4: (SupplO73) Pessi T, Koivula T, Kaukinen S, Marnela K M 1985 Significance of branched chain amino acids in parenteral nutrition of patients with severe acute haemorrhagic pancreatitiss. Clinical Nutrition 4: (SupplO74) Bonau R A, Daly J M, Moldawer L, Blackbum G 1985 Muscle amino acid flux in patients receiving BCAA solutions. Journal of Parenteral and Enteral Nutrition 9: 66 Bastow MD 1986 The complications of enteral nutrition. Gut 27: supplement I 51-55 Keohane P I’, Attrill H, Jones B J M, Silk D B A 1983 Limitations and drawbacks of ‘fine bore’ nasogastric feeding tubes. Clinical Nutrition 2: 85-86 Dobbie R I?, Hoffmeister J A 1976 Continuous pumptube enteral hyperalimentation. Surgery Gynaecology and Obstetricss 143: 273-276 Rombeau J L, Jacobs D 0 1984 Nasoenteric tube feeding. In: Enteral and Tube Feeding. J L Rombeau, MD Caldwell (eds) WB Saunders Co, Philadelphia pp 261-274 Keohane P P, Attrill H, Silk D B A 1986 Clinical effectiveness of weighted and unweighted ‘fine bore’ nasogastric feeding tubes in enteral nutrition: a controlled clinical trial Journal of Clinical Nutrition Sr Gastroenterology 1: 189-193 Rees R G, Attrill H A, Quinn D G, Silk D B A 1986 Improved design of nasogastric feeding tubes. Clinical Nutrition 5: 203-207 Twomey P L, St John J 1984 The neurologic patient, In Enteral Tube Feeding. J L Rombeau, MD Caldwell (eds) WB Saunders Co, Philadelphia pp 293-302 Casewell M W 1979 Nasogastric feeds as a source of Klebsiella infection for intensive care patients. Research and Clinical Forums 1: 101-106 Anderton A 1983 Microbiological aspects of the preparation and administration of naso-gastric and naso-enteric tube feeds in hospitals-A review. Human Nutrition Applied Nutrition 37A: 426-440 Bastow MD, Greaves P, Allison S P 1982 Microbial contamination of enteral feeds, Human Nutrition Applied Nutrition 36A: 213-217

Submission date: 4 February

1987

[891 Rees R G I’, Ryan J, Attrill H A, Silk D B A Clinical evaluation of 2 L pre-packaged enteral diet delivery system-A controlled trial. (submitted to JPEN) [901 Silk D B A 1986 Enteral nutrition: The future. Journal of Clinical Nutrition and Gastroenterology 1: 91-96 1911 Spiller R C 1985 The influence of fat on human small bowel motility. Cambridge: University of Cambridge. (Thesis) [921 Olivares L, Segoria A, Refuella J 1974 Tube feeding and lethal aspiration in neurological patients: a review of 720 autopsy cases. Stroke 5: 654-657 [931 Kiver K F, Hayess D I’, Fortin D F, Main B S 1984 Pre- and post-pyloric enteral feeding: analysis of safety and complications. Journal of Parenteral and Enteral Nutrition 8: 95 [941 Rombeau J L, Barot L R, Low D W, Twomey P L. In: Clinical Nutrition Vol I. Enteral and tube feeding. J L Rombeau, M D Caldwell (eds) WB Saunders Co, Philadelphia pp 275-291 I951 Sriram K, Peterson J, O’Gara J et al 1986 The continent gastrostomy (abstract). Journal of Parenteral and Enteral Nutrition 10: 21 S [961 Snape W J, Battle W M, Schwartz S S et al 1982 Metoclopramide treatment of gastroparesis due to diabetes mellitus. Annals of Internal Medicine 96: 44P 446 I971 Kirby D F, Craig R M, Tsang T K, Plotnick B H 1986 Percutaneouss endoscopic gastrostomies. A prospective evaluation and review of the literature. Journal of Parenteral and Enteral Nutrition 10: 155-159 [981 Rees R G P, Keohane P P, Grimble G K, Frost P G, Attrill H, Silk D B A (1985) Tolerance of an elemental diet administered without starter regimen. British Medical Journal 290: 1869-1879 I991 Spiller R C, Higgins B E, Frost P G, Silk D B A 1984 Inhibition of jejunal water and electrolyte absorption by therapeutic doses of clindamycin in man. Clinical Science 67: 117-120 WOI McNeil N I, Cummings J H, James W P T 1978 Short chain fatty acid absorption by the human large intestine. Gut 19: 819-822 WI Ruppin H, Barr-Meir S, Soergel K H, Wood C M, Schmitt M G 1980 Absorption of short chain fatty acids in the colon. Gastroenterology 78: 1500-1507 w21 Patil D H, Westaby D, Mahida Y R, Palmer K R, Rees R G, Clark M L, Dawson A M, Silk D B A Comparative modes of action of lactitol and lactulose in the treatment of hepatic encephalopathy. Gut (in press)