Discussion leader: R. A. Argenzio
THE ROLE OF DIETARY FIBRE IN DIGESTION ABSORPTION AND METABOLISM A.G. LOW
National Institute for Research in Dairying, Shinfield, Reading, Berkshire, RG2 9AT, England*
SUMMARY
Dietary fibre is composed of a very heterogenous mixture of substances, mainly associated with plant cell walls, which may be defined as non-starch polysaccharides and lignin. The complexity of its physical and chemical properties make accurate and
reproducible detailed analysis very difficult. Raising the dietary fibre content of the diet tends to increase voluntary feed intake.
Gastric, pancreatic and small intestinal secretions are generally higher when additional dietary fibre is given: in particular nitrogenous secretions increase and this is a factor determining the apparent digestibility of proteins. Transit is often faster when dietary fibre is added to low fibre diets but little effect can be seen in cereal-based diets. In general the rate of nutrient absorption is decreased by additional dietary fibre: in addition the amount of energy absorbed in the small intestine is usually reduced, to a greater extent than for lipids and
nitrogenous compounds. Fermentation of dietary fibre in the large intestine increases with aqe. The products (mainly volatile fatty acids) are absorbed and metabolized but reliable guantitative estimates of their contribution to energy supply are lacking.
INTRODUCTION
In recent years there has been renewed interest in the role of dietary fibre in pig nutrition. This arises from the increasing pressure placed upon pig producers to find alternative and cheaper types of feedstuffs, in view of the widespread demands that grains and protein concentrates should only be eaten directly by man.
* Address from 1 April 1985: The Animal and Grassland Research Institute, Shinfield, Reading, Berkshire, RG2 9AT, England.
Durinq the last twenty vears or so it has become clear from epidemioloqical studies of human populations that there is an apparent inverse relationship between the consumption of a diet that is rich in foods which contain cell walls (e.g. high
extraction cereals, fruit, vegetables) and the incidence of such diseases as diabetes, diverticular disease, large bowel cancer, coronary heart disease, gallstones and obesity. This evidence is often described in terms of the "dietary fibre hypothesis". This hypothesis also includes the statement that a diet providing a low
intake of cell walls is a causative factor in the aetiology of the disease, in some cases, while in others it provides conditions under which other factors may be more active (Southgate, 1982).
Interest in the first of these issues has led recently to an increase in research on the role that dietary fibre can play in pig nutrition; the second issue has led to the recognition that the digestive system and metabolism of pigs are strikingly similar to those of man and that the responses of pigs and man to some types of dietary fibre are similar (Leeds, Kang, Low and Sambrook, 1980). Thus the pig appears to be a good model for man in this context. This review will include discussion of the role of dietary fibre both in agricultural aspects of pig nutrition and of studies in pigs which may help to explain the mode of action of dietary fibre in man.
DEFINITIONS
Much has been written on this topic and it is clear that there is no complete definition of the term dietary fibre. Indeed the definition depends upon the viewpoint of the person attempting to state what dietary fibre is. However, two definitions appears to be useful: (a) "the sum of the polysaccharides and lignin which are not digested by the endogenous secretions of the digestive tract" (Trowell, Southgate, Wolever, Leeds, Gassull and Jenkins,
1976) and (b) "non starch polysaccharides and lignin" (Southqate, 1982). The first is a conceptual definition embracing chemical, physical and physiological aspects of fibre which cannot yet be measured fully, while the second describes an entity that can be measured using existing methods, though not without difficulty.
It is important to recognize that these definitions include not only the constituents of plant cell walls, but also non-starch polysaccharide storage gums, extracted from plants such as guar gum, which do not fulfill a function within the cell wall, and algal polysaccarides such as carrageenans.
CHEMICAL AND PHYSICAL PROPERTIES
Apart from lignin, which is an aromatic polymer of phenolic alcohols, dietary fibre consists of a very wide range of polymers of pentoses (especially arabinose and xylose) and hexoses
(especially glucose, fructose, galactose). The principal structural group include the ß-glucans (cellulose and ß-(l-3, l-4)-glucans and the heteroglycans (pectic substances,
hemicelluloses, storage and exudate gums and algal polysaccharides such as agar, caraqeenans and alginates). In addition the
oligosaccharides raffinose and stachyose, which are found in legumes, are components of dietary fibre. It has also been proposed that certain types of processed starch which contain ester, ether or phosphate derivatives ('resistant starch') and which provide steric blocks to host amylase activity should also be reqarded as dietary fibre. A detailed account of the chemistry of dietary fibre was presented by Southqate (1976).
The ability of dietary fibre to hold water in varying amounts has often been used as an explanation for differing effects of dietary fibre on faecal output. The water holding capacity of dietary fibre i_n vitro varies enormously depending on source, maturity of the plant, whether or not it has been processed, particle size, the pH and electrolyte composition of the
surrounding fluid and so on. Nevertheless, Stephen and Cummings (1979) found that those dietary fibres which hold the least water
in vitro (e.g. bran) produced the largest increases in faecal output, while gums such as pectin which hold large amounts of water jjn vitro produced almost no change in faecal output.
Increases in faecal water output following dietary fibre
supplementation are also the result of larger bacterial numbers (Stephen and Cummings, 1980); bacteria typically contain 80%
water. It is generally found that coarsely ground dietary fibre leads to a greater water-holding capacity than when it is finely ground and that it induces a greater faecal output. Dietary fibres also tend to have ion exchange properties for most monovalent cations (McConnell, Eastwood and Mitchell, 1974) and calcium (James, Branch and Southgate, 1 9 7 8 ) . In addition
pH-dependent adsorption onto dietary fibre is known to occur, for example in the case of bile acids and seems to be associated with suppression of ionization within the dietary fibre (Eastwood and Hamilton, 1 9 6 8 ) . Thus dietary fibre includes a complex variety of chemical structures together with a wide spectrum of physical properties, which give rise to diverse physiological and
nutritional effects. There is inadequate detailed understanding of the relationship between the structure and function of dietary fibre and this is a barrier to progress in its use both in animal nutrit-ion and medicine. It is important to recognize the
complexity and diversity of the mechanisms involved: it is evident that many types of dietary fibre can influence the intake,
digestion, absorption and metabolism of all the major nutrient types.
ANALYTICAL METHODS
At a practical level dietary fibre is defined by the analytical method used for its measurement. A wide range of methods is available but comparisons between them show major differences, because of the differing susceptibility of the complex mixture of substances to different extraction procedures.
Crude Fibre This method was developed over 150 years ago as a means of measuring the indigestible fraction of animal feedstuffs.
The sample is treated sequentially with petroleum ether, hot sulphuric acid, boiling water and alkali. The resultant insoluble residue contains mainly cellulose and lignin but recovery is rarely complete (Van Soest and McQueen, 1 9 7 3 ) . Nevertheless this remains a standard method in many countries.
Neutral Detergent Fibre This method, described by Van Soest and Wine (1967) involves digestion of the sample in neutral detergent solution. After filtration the residue is dried and weighed.
Although lignin and cellulose are usually fully recovered,
hemicellulose recovery may be incomplete because its water-soluble components are largely lost during filtration.
Acid Detergent Fibre This method compliments the previous
procedure and was developed by Van Soest (1963). Digestion of the sample in acid detergent solution is followed by filtration, and the residue is then dried and weighed, to provide a measure of the cellulose and lignin contents of the sample. Other components of dietary fibre are largely excluded.
Non Starch Polysaccharide Methods Initial removal of starch (by enzymic hydrolysis) is followed by separation into cellulose, non-cellulose polysaccharides and lignin, acid hydrolysis and finally colorimetric (Southgate, 1969) or gas liquid
Chromatographie (Englyst, Wiggins and Cummings, 1982) measurement of the individual sugar constituents of each fraction. A rapid enzymic procedure has recently been described by Asp, Johansson, Hallmer and Siljestrom (1983): following this hydrolysis samples can be analysed in as much detail as reguired.
Although the crude fibre, acid and neutral detergent fibre methods of analysis continue to have value in practical pig
nutrition because of the good inverse correlation between the values obtained and the digestible or metabolizable energy content of the diet, they do not provide sufficient information for
elucidating the detailed effects of dietary fibre. The importance of accurate and detailed chemical and physical characterization of dietary fibre in experimental work cannot be over-emphasised. In addition agreement on a standard method of analysis would
transform our ability to interpret results from different research centres. It is apparent from a number of comparative studies that different methods of analysis give very different results for components of dietary fibre; for example, Millard and Chesson
(1984) prepared dietary fibre extracts from swede (Brassica napus) as fed and from the ileal digesta of pigs by six different
published methods. The individual components of both the insoluble and soluble fractions were then analysed by a single method to give remarkably different results in many instances;
examples of the range of values in the insoluble fraction (g/kg feed) are: arabinose 4.5 0.2, xylose 8.4 2.1, galactose 7.0 -0.0, phenolics 4.4 - 2.0. The lowest values in each case were obtained using the acid detergent fibre method and the high values
from recently developed extraction methods. A comprehensive review of analytical procedures for dietary fibre has been edited by James and Theander (1981).
DIETARY FIBRE AND FEED INTAKE
It is well established that as the dietary fibre content of the diet increases, so the voluntary feed intake of pigs
increases: the ARC (1967) concluded that for every 1% increase in dietary fibre content up to a total of 100g/kg diet (measured as crude fibre) a 3% increase in diet intake occurred. At the same time the growth rate falls, because increased appetite does not compensate fully for the reduced dietary energy concentration as the dietary fibre content increased particularly in young pigs (Owen and Ridqman, 1968). Furthermore the carcass weight as a percentage of liveweight falls as the dietary fibre content of the diet is increased, because of the larger weight of gut contents and the heavier gut tissue. Behind such general statements lies much uncertainty about the responses of pigs to specific types of dietary fibre: no thorough comparison of the palatability of different types of dietary fibre is available. When the dietary fibre content of the diet was increased beyond 100g crude fibre/kg diet, as in the studies by Baker, Becker, Jensen and Harrison
(1968) on growing pigs given 0-400g cellulose/kg diet, then the
voluntary feed intake fell from 2.63 to 1.50kg/day, and daily gain fell from 760 to 250g/day.
Zoiopoulos, English and Topps (1982) found that the voluntary feed intake of sows was 7.79kg/d (85.0 MJ digestible energy) when given a diet containing 400g/kg of oat husks; corresponding values of 5.80kg/d (60.4 MJ digestible energy) were observed when the diet contained 300g/kg barley straw. At present it is not known what attribute of these two sources of dietary fibre led to the differences in feed intake. Information on the responses to different types of dietary fibre is of particular importance when attempts are made to control the feed intake of breeding animals, in which excessive gains can occur under a_d_ libitum systems. In the same way, knowledge of the effects of specific attributes of different dietary fibres on intake and satiety would be of great value in clinical nutrition.
An interesting hypothesis to explain how moderate intakes of dietary fibre may increase voluntary feed intake has been proposed by Bergner (1981). The bacterial flora of the large intestine hydrolyses undigested proteins and transforms the amino acids so released to tyramine and tryptamine, amine derivatives of tyrosine and tryphophan, respectively. These can saturate the satiety centre of the hypothalamins and reduce feed intake. However, by lowering the pH in the gut lumen, by increased production of volatile fatty acids from diets with a higher dietary fibre
content, this effect could be reduced, because the am i ne-producing bacteria are only active at a relatively high pH.
EFFECTS OF DIETARY FIBRE ON DIGESTIVE SECRETIONS
The effects of dietary fibre on digestive secretions in pigs appear to be considerable. For example, mean volumes (litres) of saliva and gastric juice, bile and pancreatic juice for low and high-dietary fibre diets (based on (a) starch, casein and
cellulose or (b) barley and either fishmeal or soyabean meal) secreted per 24h in 40kg pigs were respectively: 4.0, 8.0 (saliva and gastric juice), 1.2, 1.7 (bile), 1.2, 2.2 (pancreatic juice)
li*
(Zebrowska, Low and Zebrowska, 1983; Sambrook, 1981): the neutral detergent fibre contents of the diets were 50g/kg and 180g/kg respectively, while crude fibre concentrations were similar, emphasising the large hemicellulose content of the barley-based diet. While these results should not be taken to indicate a definite link between dietary fibre and secretion, by far the largest difference between the diets was in their dietary fibre content. The reasons for such apparent effects remain to be eluc idated.
Certain types of dietary fibre increase the viscosity of meals and of the gut contents; one example is guar gum, which has been found to increase nitrogen secretion in isolated loops of jejunum in conscious growing pigs from 35 to 67 mg/m/h (equivalent to an increase from 15 to 25g/d if secretion occurs at a uniform rate throughout the small intestine (Low & Rainbird, 1984). The increased nitrogen contains both protein and DNA; the latter is likely to be a constituent of mucosal cells, the exfoliation of which increased following guar gum addition to diets in rats
(Johnson, Gee and Mahoney, 1984). Furthermore intestinal protein synthesis in rats was increased when they consumed 99g dietary fibre/kg from a cereal based diet rather than a semi-synthetic diet containing 40g cellulose/kg (Southon, Livesey, Gee and
Johnson, 1 9 8 5 ) . A similar effect has been observed on the flow of nitrogen and amino acids passing through the ileum of pigs
receiving protein-free diets with supplementary cellulose by Sauer, Stothers and Parker (1977) and Taverner, Hume and Farrell (1981). Cellulose supplementation of diets has also been shown to increase the output of faecal nitrogen, to a greater extent than oat hulls, while methylcellulose had little effect (Whiting and Bezeau, 1 9 5 7 ) . These results have important implications for interpretation of data on the apparent digestibility of nitrogen:
it seems possible that as both insoluble and soluble types of dietary fibre can enhance endogenous nitrogen secretion, so the apparent digestibility of nitrogen may be a function not only of
the inherent digestibility of the dietary protein, but also of the type and amount of dietary fibre with which it is incorporated in the diet. Further information on this topic could have important implications in relation to the efficiency of protein digestion.
Dietary fibre may also influence the movement of water in the gut: for example the rate of gastric emptying of water is reduced by guar gum (Rainbird and Low, 1983) while additional dietary cellulose greatly increases the volume of digesta in the ileum and faeces of pigs (Partridge, 1978). Similar effects of dietary fibre on gut water volume and faecal output have been observed by other authors, but so far it has not been possible to identify the features of the fibre concerned. It has rarely been possible to demonstrate whether inhibited water absorption or increased secretion is occuring, but Rainbird, Low and Zebrowska (1984) found that guar gum decreased the net absorption of water in isolated loops of jejunum in pigs.
EFFECTS OF DIETARY FIBRE ON GUT MOTILITY AND TRANSIT So far there is very little information on the effects of dietary fibre on motor activity in either the small or large
intestines of pigs. Sissons, Rainbird and Thurston (1984) showed that gastric motility appeared to be unaffected by addition of guar gum to the diet, but duodenal motility was reduced when assessed in terms of the duration of periods of irregular spike activity (Sissons and Rainbird; unpublished results). Changes in motility of this kind can be expected to modify the time course of glucose absorption (Rayner, Gregory and Goodall, 1984).
Bran has been shown to stimulate propulsive colonie motility in pigs (Fioramonti and Bueno, 1980); this appears to be the result of poorly understood direct mechanical factors rather than of fermentation products, including VFA (Bardon and Fioramonti 1 9 8 3 ) . This is an area of gastrointestinal physiology that is developing rapidly and it is likely that studies with pigs can contribute to basic understanding of the effects of dietary fibre on the digestive processes.
Although the effects of dietary fibre on transit of dietary components are doubtless mediated in part by motility phenomena, this topic has hardly been explored in any species. Studies in man on overall transit time have often been made on the assumption
that this is a physiologically significant measure of gut
function, but this has not been clearly demonstrated. It is now thought that transit may influence rather than be the result of events related to diet type in the large intestine (in which dietary residues spend most of their total transit time). For example Cummings, Southgate, Branch, Houston, Jenkins and James (1978) found that faeces weight varied from 65 to 194g/d in 16 normal subjects all eating the same amount of dietary fibre, while transit time ranged from 31 to 117h; the length of transit time varied inversely with faecal weight. Further evidence that transit time can alter faecal output comes from studies in man by Stephen (1980) who found that pharmacological slowing of transit was accompanied by a fall in faecal output, and vice versa ; in the same studies it was also found that there is an inverse
relationship between the extent of bacterial degradation of dietary fibre and transit time. Thus in the case of man substantial individual differences in transit time of the same diet are one aspect of the function of the large intestine, which can also be modified by such factors as the fibre source and the physical form of the diet. It is evident that there is also considerable variation in the digestibility of dietary fibre in different pigs and it seems likely that here too individual
transit times are found, but as in man the reasons for this remain unknown.
Th e results of some recent studies on the effect of dietary fibre on overall transit time in pigs are summarised below.
Table 1. Effects of dietary fibre on overall mean transit time in pigs (Data from: Kuan, Stanogias and Dunkin (1983) ( 1 ) , Ehle Jeraci, Robertson and Van Soest (1982) ( 2 ) , Fioramonti and Bueno
(1980) ( 3 ) , Bardon and Fioramonti (1983) ( 4 ) , and Canguilhem and
It is evident that additions of various types of dietary fibre can have marked effects upon transit time of diets, especially if
It is evident that additions of various types of dietary fibre can have marked effects upon transit time of diets, especially if