Dietary Fibre: Nature’s Elixir or Snake Oil

A critical analysis of the data on prophylactic action of dietary fibre

against Colorectal Cancer

Robin L. M. Cheung

Student #244679

For Dr. Steven Brooks

Health Protection Branch,

Health Canada

61.491* Special Directed Studies

Carleton University

Introduction

The bane of Western culture, colorectal cancer is the second most common cause of cancer-related deaths, abdicating only to lung cancer. It is a disease largely modulated by environmental factors, and much current belief centres around the notion that dietary fibre may play a role in the prevention of colorectal cancer. To this end, many years and many studies have been dedicated to the elucidation of mechanisms involved in the commencement and development of colonic tumours, that they might provide valuable insights into putative treatments and preventive measures that can be taken to help reduce the cancer rate. From the time that Burkitt (1970) first posited that dietary fibre possessed cancer protective, cancer-preventive properties, a great number of studies have been conducted, with equivocal results. While the precise mechanisms involved have not yet been elucidated, it is apparent that populations with fibre-rich diets generally exhibit the lowest rates of colorectal cancer.

Alabaster et al. (1996) indicate that colorectal cancer in humans is preceded in 50% of tumours by a point mutation in the gene, K-ras. It is considered to be a genetic hallmark signalling increased risk of developing colorectal cancer (Sills et al., 1999). K-ras mutations are reliable indicators for cancer development potential in that point mutations at these genes greatly increase the risk of developing cancer in a significant number of cases. Kubrusly et al. (1999), for example, cite a 90% observation rate in the case of pancreatic adenocarcinomas. Sills et al. (1999) illustrate the importance of the K-ras gene in mice using 1,3-butadiene, a potent carcinogen. They remarked an 80% mutation rate in K-ras genes in lung neoplasms compared with a latent 30% background mutation rate.

That colonic cells already experience some of the highest turnover rates in the human body (Wasan and Goodlad, 1996) itself only facilitates cancer development. In fact, Wasan and Goodlad (1996) propose that this hyperproliferation, itself, is an important early stage in the development of cancer–something already manifested in the metabolic processes of colonic cells.

Definition of Dietary Fibre

Long the subject of debate, there has never been a consensus regarding a precise definition of “dietary fibre.” Indeed, the term itself is a misnomer–many components included in this family are fibrous at all. There are, however, certain broad definitions which upon which many classification schemes have been based.

Dietary fibre has been defined to include components of plant cell walls and the components of these cell walls (Harris and Ferguson, 1999), such as cellulose and other structural components. Indeed, however, this classification necessarily includes cell wall components and their derivatives, such as pectins and carboxymethyl cellulose, a chemically-modified cell wall constituent. While one of the basic tenets of the classification schemes used to identify dietary fibres has been that the substance be undigestible by the human alimentary tract, this criterion is in itself a source of great variability and debate, since materials may be undigestible by enzymes produced by the human body, yet digestible by indigenous microorganisms which inhabit the gut. Indeed, in Western diets, plant cell walls contribute roughly 95% of the daily intake of dietary fibres (Harris and Ferguson, 1999). Classically, non-starch polysaccharides from sources other than plants have also been included in this definition, such as those originating from micro-organisms, seaweeds, and exudates. The inclusion of such materials as gums and mucilages has further confounded the nomenclature of such materials, as these substances can no longer be characterized as being fibrous, either structurally or molecularly.

The introduction of non-fibrous members to the “dietary fibre” family naturally led to the subclassification of the family into two still broad species, insoluble fibres and soluble fibres. The determination of whether a compound is considered insoluble or soluble has been determined analytically by in vitro studies of solubility in water or other buffer solutions. This, in itself, lends great variability and debate as to whether a component can be classified as insoluble or soluble, and is highly dependent upon the methodology used.

Cellulose is one of the predominant plant cell wall-derived compounds considered dietary fibre. It is a linear molecule consisting of 1,4-linked beta-glucosyl residues (Harris and Ferguson, 1999). Cellulose microfibrils are roughly 3-10nm in diameter, plant species dependent. While vegetable or fruit cell walls may comprise 30-40% cellulose, cereals, known for their lignin content, may contain only up to 4% cellulose (Harris and Ferguson, 1999). It is not generally fermented by either human endogenous digestive enzymes or intestinal microflora present in the human alimentary tract (Wijnands et al., 1999).

Pectins represent a complex family of soluble fibres that comprise a diverse array of polysaccharide domains known as “homogalacturonan” domains (Jarvis, 1984). Parenchymatous cells of fruits and vegetables are known to contain high concentrations of pectins, but cereals, as with cellulose, generally contain only very small amounts of pectins. There is evidence that pectins may play a protective role against cardiovascular disease by interfering with cholesterol metabolism (Schneeman, 1999; Brown et al., 1999).

Hemicellulose comprises a family of polysaccharides extracted from cell walls with alkali (Harris and Ferguson, 1999). Commonly found in cereal cell walls as arabinoxylans (1,3- and 1,4-) beta-glucan linked sugars on a back-bone of xylosyl residues substituted with arabinosyl residues. Ferulic acid, commonly known to inhibit the production of flatulent gasses, is often ester-linked to some of the arabinose residues.

There have been several methodologies proposed to determine the solubility and content of dietary fibres in various foodstuffs. Historically, fibre was assayed as a representation of the crude fibre content. This represents the amount of material remaining after a simple chemical acid-alkali treatment which was designed to approximate the environmental conditions in the human digestive system through to which the food would be subjected (Van Soest, 1978). It lacks the ability to account for materials which would be digested by enzymatic methods. More importantly, however, it tends to cause the decomposition of materials which would normally pass through the digestive system largely undigested because it is a brute force approach (Slavin, 1987). This method, while retained for use in animal foodstuffs crude fibre assays, has largely been abandoned as a means of estimating dietary fibres in the human foods industry due to its gross margin of intrinsic error. Slavin (1987) estimates that crude dietary fibre estimates often underestimate the dietary fibre content of foods three- to five-times lower than actual physiological values; further, there can be no consistent adjustment factor due to account for the activities of digestive enzymes and intestinal microflora. Further, there are proteinaceous components in foods that, in the absence of human digestive enzymes and physiological environments may be inaccurately included in a dietary fibre assay. Slavin (1983) in an earlier paper cite the example that while commercially-prepared bran flakes contain six times more dietary fibre than a crude acid/alkali assay would report, strawberries contain only 1.6 times more. While the relationship is consistently one of underestimation, because there is no clear-cut method to establish a correction factor, crude fibre estimates are largely inaccurate and insufficient for the stringent demands of regulatory bodies for purposes such as labeling, and grossly inadequate for purposes of establishing a relationship between dietary fibre intakes and health effects. Consequently, more elaborate methodologies have been proposed to approximate more accurately the dietary fibre content in foodstuffs.

One of the more common approaches to a more accurate determination of fibre content revolves around the Van Soest detergent method. One such methodology is the Neutral Detergent Fibre (NDF) method. These are simple, one-step procedures that rely on detergent action to exclude cytoplasmic components and water-soluble plant cell wall materials. Inasmuch as the procedure yields estimates of insoluble plant material, it underestimates soluble fibre components, such as pectin (Ross et al., 1985). Although this method reports only insoluble fibre content, many handbooks incorporate these values, in conjunction with separately-reported soluble fibre contents to represent estimates of published dietary fibre contents. To address this issue, a refinement of the NDF method, an enzyme-modified NDF procedure, described by Heckman and Lane (1981) was proposed to incorporate the enzymatic actions to which foods are exposed in the digestive systems. This system, while more accurate in its incorporation of enzymatic effects, still lacks the ability to account for water soluble fibres.

In order to address the issue of differential soluble-insoluble fibre extractions in the NDF methodologies, Hellendoorn et al. (1975) proposed a single-fraction total dietary fibre method designed to yield a single-fraction total dietary fibre estimate. Differing fundamentally to the crude fibre acid/alkali assay, the Hellendoorn method relies on the use of mammalian enzymes to digest the protein and starch components of foods, a method closer approximating physiological conditions. Although the Hellendoorn method fails to estimate water soluble fibres, when used in conjunction with gravimetric techniques, it can often report estimates of water soluble and water insoluble fibres.

The Southgate procedure was outlined in 1969 and is still in use for analytical dietary fibre assays. In fact, many current dietary fibre estimates were prepared using the Southgate procedure. While a much more complex procedure than the Hellendoorn or NDF methods, it yields a much more accurate representation of the dietary fibre content in foods. Southgate (1969) described a procedure where non-fibrous components of foods are removed sequentially by a combination of chemical and enzymatic means, yielding total dietary fibre content. The southgate procedure relies on gravimetric techniques to extract hemicellulose, pectin, cellulose, and lignin. Although certain starches are often included in this estimate, Stephan et al. (1983) point out that significant amounts of starch pass through much of the human gut undigested, some of these starches reported as “dietary fibre” by the Southgate procedure represent an estimate of the starches that escape digestion in physiological conditions and, in fact, may function as dietary fibre in many respects.

A simplification of the Southgate procedure, yielding still physiologically-relevant estimates of dietary fibre content in foods is a methodology proposed by Furda (1977). The Furda method represents a two-fraction refinement of the Southgate procedure, simpler analytically and practically. Heckman and Lane (1981) recommend the use of the Furda method on the basis of this practical refinement.

COMPONENTS OF DIETARY FIBRE: What is assayed in these measurements?

Scheeman (1986) point out that the various molecular components of dietary fibre can be classified in three major fractions:

Structural Polysaccharides

Cell wall components

Non-cellulose polysaccharides

· hemicellulose

· pectins

Cellulose

Structural Non-Polysaccharides

Lignin

Non-structural Polysaccharides

Gums, Mucilages

Algae and seaweed polysaccharides

According to the definition by Slavin (1987), dietary fibre comprises the portion of plant cells that cannot be digested by human alimentary enzymes and is therefore not absorbed by small bowels. While only one component of cell walls, cellulose, is truly fibrous, this misnomer has been expanded to include the compounds listed below:

TABLE 1 CLASSIFICATION OF CHEMICAL COMPONENTS OF DIETARY FIBRE

Table 1. Classification of chemical components of dietary fibre

Fibre

Chemical Components

Polysaccharide

Main Chain

Side Chain

Description

Cellulose

glucose

none

main structural component of plant cell walls

Non-cellulose

Hemicellulose

xylose

mannose

glucose

arabinose

galactose

cell wall polysaccharide

backbone of 1-4-linked pyranoside sugars

Pectic Substances

galacturonic acid

rhamnose

arabinose

xylose

fucose

galactose

part of primary cell wall and middle lamella

Mucilages

galactose-mannose

glucose-mannose

arabinose-xylose

galacturonic acid-rhamnose

galactose

synthesized by plant secretory cells

used in food industry

Gums

galactose

glucuronic acid-mannose

galacturonic acid-rhamnose

xylose

secreted at site of plant injury

specialized secretory cells

Algal polysaccharides

mannose

xylose

glucuronic acid

glucose

galactose

derived from algae and seaweed

Non-polysaccharide

sinapyl alc.

coniferyl alc.

p-coumaryl alc.

3-dimensional structure

non-carbohydrate

phenyl propane polymer

Most starches are easily digested within the human alimentary canal by endogenous enzymes (Slavin and Ferguson, 1987). It is natural, however, for some starches to escape digestion due to their physicochemical properties. Englyst et al. (1992) describe three types of resistant starches which routinely pass through the small intestine. The first are starch granules which are physically contained within plant cells. While they may be degraded by endogenous enzymes, they are shielded from enzymatic action physically by being contained within cells. There are also starches which prove resistant to many carbohydrate-degrading enzymes, such as alpha amylase, including some starches found in the potato. The third type of resistant starch is found in cooked and processed foods, such as breads or cooked potatoes, and represents a “retrograded” starch polymer.

HEALTH BENEFITS OF DIETARY FIBRE

Dietary fibre has long been known to be a cure for constipation, ever since the ancient classical civilizations. Although the term was formally defined first only in 1953, Burkitt (1969) is generally credited with being the first to propose that dietary fibre might have a protective effect against many western ailments. Among the beneficial protective effects dietary fibre is credited with providing are those relating to diabetes, cardiovascular disease, and most commonly, cancers such as colorectal cancer and breast cancer. Since the initial proposition, many studies have been conducted to substantiate or refute these claims, many of which yield equivocal results, likely due to substantial differences in experimental methodologies.

CARDIOVASCULAR

Foods high in dietary fibre have been proposed to have a protective effect against cardiovascular disease (Schneeman, 1999). Recent studies supporting this claim seem to reach a general consensus that in order to provide a protective effect against cardiovascular disease, dietary fibre must be of the soluble type; insoluble fibres, such as wheat bran fibre (largely lignin) provide little protective effects, exerting any such effects only through the dietary displacement of foods that supply dietary fats or cholesterols (Brown et al. 1999). The Brown et al. (1999) study proposed that while moderate intakes of soluble dietary fibre, such as pectin or gums are able to lower serum cholesterol, which is generally interpreted to be beneficial. Further, Brown et al. (1999) pointed out that while soluble fibres were the only type able to effect lowering of serum cholesterol, soluble psyllium fibres and guar gum effected significant, though still minimal, lowering of HDL cholesterol, which is not necessarily interpretable as a beneficial effect with respect to cardiovascular disease protection. This further supports the notion that a simple contrived classification of “soluble” or “insoluble” is neither physiologically nor analytically relevant.

DIABETES

Anderson (1992) showed that high fibre diets were able to decrease post-prandial glycaemia, which is an important factor when considering diabetic influences. Traditionally, this has been attributed to interference to nutrient uptake in the gastro-intestinal tract. Bakker et al. (1998) point out, however, two important factors when considering the proposition that high fibre diets are inversely correlated with blood glucose levels after feeding. Three key enzymes involved in glucose metabolism are pyrvuate dehydrogenase, alpha-ketoglutarate dehydrogenase, and transketolase. It is known that thiamine is an important cofactor in the optimal activity of these enzymes, and that thiamine deficiencies result in an overall reduction in the production of insulin by beta cells and in the oxidation of glucose and the citric acid cycle. An impairment in these pathways, possibly due to low thiamine levels, could result in the efficient glucose processing evidenced by the lack of thiamine in a low fibre diet. While a deficiency does not imply increased efficiency when in abundance, it is useful to note that thiamine is particularly abundant in high fibre foods. The second observation on the part of Bakker et al. (1998) was that when corrected for fasting glucose, the inverse association between fibre intake and two-hour blood glucose seemed to be nullified. They proposed that this inverse correlation be attributed not to dietary fibre intake but rather to improved insulin sensitivity after fasting. If glucose tolerance is mediated by thiamine content in high-fibre foods, then this should be readily tested by a cohort study utilizing thiamine supplementation in the absence of varied fibre diets. Further, thiamine is abundant in such food products as whole grains, unprocessed rice, and legumes. A dietary cohort study involving these food sources may also shed insight into the significance of thiamine with respect to insulin activity and glucose tolerance.

It is possible also that dietary fibre is able to mediate glucose metabolism by diluting ingested carbohydrate and thereby moderating the absorption and metabolism of glucose. Further, undigested fibres may be able to displace digestible carbohydrates in a diet, also affecting blood glucose levels.

COLORECTAL CANCER

Nearly a century ago, it was proposed that the anaerobic microflora in the gut were responsible, through their production of waste products from methabolism of nitrogenous compounds, such as proteins and hetrocyclic amines, for toxification and, in fact, ageing. At around the same time, the idea was proposed that carbohydrate metabolism by lactic acid bacteria contributed to a favourable colonic environment (Hill, 1998).

Conventional wisdom surrounding protective benefits of dietary fibre against colorectal carcinogenesis seems to be centered upon four main tenets (Burkitt, 1970): 1. Stool bulking, diluting carcinogens in the organism’s food source (Cummings et al., 1992); 2. Fermentation of carbohydrate, specifically dietary fibre components in the food source; 3. Changes in intestinal microflora brought about my dietary fibre as a micro-organism’s food-source and favourable environment; and 4. Modificiations to the physiological environment of the intestinal lumen, brought about by the above factors. More recent studies (Harris et al., 1991) propose mechanisms of protection that involve hydrophobic interactions between carcinogenic compounds and hydrophobic components of insoluble dietary fibres, such as lignin. Whatever proposed mechanisms may be involved, the data to date are equivocal regarding the specific components involved or mechanisms involved in the protection against colorectal cancer by dietary fibres. It is immediately apparent from the literature that the interactions between dietary fibre, intestinal microflora, and the colonic environment and metabolic processes are complex and this is borne by the data to date. There have been many studies both supporting and refuting dietary fibre as a nutritional approach to cancer prevention. Some studies even support the conservative view that dietary fibre supplementation may be detrimental to health (Wasan and Goodlad, 1996) or even promote development of tumours (Harris et al., 1999). There are many factors involved in the study of dietary fibre intake with respect to colorectal cancer and prevention. There have been two main approaches taken to study these relationships. One approach is a more wholistic one, taking into account observation of overall dietary intakes, the other being a more componential approach, studying the effects of the chemical compounds considered to be dietary fibres in isolation or in groups, without considering whole diet. It is clear that a more definitive study should consider dietary fibres as nutritional constructs as well as the individual components that are considered to be dietary fibres. Yet there is plenty of data to support numerous hypotheses when these data are considered as a whole. Pareto analysis, the practice of organizing factors according to magnitude of impact and effect, can yield an overall conclusion regarding the effects of dietary fibre on colorectal cancer prevention. It is likely, therefore, that studies involving dietary fibre compounds (cell wall components) studied from a plant source and not as compounds in isolation more accurately portray the interactions in vivo because certain components may be contained within plant structures until released by degradation. A study by Alabaster et al. (1996) showed that phytic acid, previously-shown to inhibit colonic tumour development (Graf and Eaton, 1993). They conceded that they believed that dietary fibre, itself, provided more anti-cancer properties but left the claim unsubstantiated.

In fact, epidemiological and cohort studies (Caygill et al., 1998; Hill, 1998; Cummings et al., 1992) seem to indicate a general trend within the populace that individuals who maintain a certain amount of dietary fibre within their diets also maintain an overall proportion meat products and dietary fats. This is intuitively supported by the proposition that was posited by Burkitt (1970) that dietary fibre would calorically displace other high-fat or nitrogenous-based foods in the organsim’s food supply. It is also possible that within the population, individuals are aware of the media thrust towards protective effects of dietary fibres against colorectal cancers and cardiovascular disease, as well as breast cancers, and the individuals who increase their intakes of dietary fibres also consciously decrease their intakes of high-fat and meat-based foods, as a conscious lifestyle choice. Stephen and Cummings (1979) further point out that most dietary fibres are extensively degraded in the colon. Although pectins are known to increase stool bulk by sequestering and binding up to 20 times its weight in water, wheat bran, which only binds three times its weight, has a more dramatic impact practically, because most fibres are degraded significantly in the colon, whereas wheat bran, high in lignin content, is largely undigested.

One study, Gaard et al. (1996) claimed that there was no cancer protective effect from milk, calcium, fish, or dietary fibre intakes; further, there was no increase in cancer incidence in high-meat, high-fat diets. This study, however, lacked adequate controls on the dietary intakes of the participants of the study. Questionnaires were submitted to participants and interpreted by nutritionists. One problem inherent with questionnaire-based meta-analyses lies with the ambiguity of food source definitions and relies on not only accurate recall by participants, but honest accurate reporting of their daily dietary intakes. Further, the study reported the only significant relevant result was that sausages, specifically, increased chances of contracting colon cancer. This questionnaire-based approach cannot take into account such important factors as food preparation, type of dietary fibre intake (through use of ambiguous terms such as “fruits” or “vegetables”), specific amount of dietary fibre and other nutrient intake, or even brand-differences between types of “bran flakes,” which in themselves can vary widely in dietary fibre, fat, and carbohydrate content. Harris and Ferguson (1999) elaborate on the difficulties interpreting such questionnaires citing differences between types of dietary fibres; further, while Western diets comprise largely whole plant cells, the analysis presumes to fit a hypothesis to simple mixtures or combinations of dietary fibre compounds.

Hill (1998) reported that the majority of epidemiological studies, perhaps with more elaborate dietary reporting or controls, show a trend towards cereals and vegetables being strongly protective, fruits showing no protective effects, and starchy root vegetables slightly promotive of colon cancer (Caygill et al., 1998). Perhaps the most significant dietary fibre component present in vegetables is generally thought to be cellulose. Wijnands et al. (1999) showed that cellulose carried little protective effects against colorectal cancer. Considering epidemiological studies often do not consider specifically the nature or preparation of vegetables, the conclusion that vegetables confer a strong protective effect against colon cancer may be due to a wide variety of factors. It has already been shown, for example, that certain plant compounds, such as phytic acid or common antioxidants such as beta carotene or d-alpha-tocopherol are able to affect development of colon cancer (Ferguson and Harris, 1999; Alabaster et al., 1996). It is possible that stool bulking and dilution of potential carcinogens is an intuitive yet substantiated physical mechanism by which dietary fibres may protect against colon cancer. Stool transit times are decreased by increased intake of dietary fibre, which can also reduce the exposure of carcinogens and secondary bile acids in the colon (Stephen and Cummings, 1979). Ingested cellulose generally contributes no energy to the organism because it passes largely through the alimentary canal undigested, undegraded. The Wijnands et al. (1999) study showed that consumption of cellulose-based foods therefore often stimulates ingestion of larger quantities of these materials in order to satisfy caloric requirements. This could be a significant contributor to stool bulking and dilution of carcinogens in the diet, in conjunction with antioxidants present in vegetables. Mature vegetables often develop a high lignin content (Slavin and Ferguson, 1987), which would confer the same protective advantages through hydrophobic interactions with carcinogenic compounds as wheat bran. The conclusion that starchy root vegetables, such as the potato, have a slight cancer promotive effect may be due to the presence of significant amounts of starches which are degradable by endogenous digestive enzymes. This would result in the provision of an abundant energy source to the highly proliferative colonocytes, which overall could contribute slightly to the development of cancer. That the relationship is only slight supports the hypothesis that the mechanism by which a high level of degradable starches could promote tumour development simply by the abundance of an available energy source. Short-chain fatty acids, especially butyrate, has been shown to moderate and even reduce colonocyte turnover, but in the absence of such moderators, in the abundance of a readily-useable energy source, it is possible that without the benefit of stool bulking, carcinogen dilution, pH-lowering, or other preventive factor, any latent or stimulated cancers could better develop (Harris and Ferguson; 1999; Wijnands et al., 1999; Kritchevsky, 1998).

Another proposed mechanism by which dietary fibres is thought to protect against carcinogenesis is through the binding of carcinogens to undegradable dietary fibres. Media frenzy of late has placed in the spotlight a deacetylated undigestible component of cretaceous exoskeletons, chitosan, which is purported to interfere with lipid absorption, thereby reducing the amount of dietary fats absorbed from foods. Similarly, Smith-Barbaro et al. (1981) and Harris et al. (1991) propose that interactions between undegradable dietary fibre components are able to interfere with interactions between known carcinogens and the colonic mucosa.

Several heterocyclic amines, produced through processing, cooking, or even charring of meats are known to be potent carcinogens. Among these heterocyclic amines known to have cancer-promotive effects are benzo[a]pyrene, 1,2-dimethylhydrazine, 1,8-dinitropyrene, and the non-aromatic amine N-nitroso-N-methylurea (Ferguson and Harris, 1996). These heterocyclic amines are largely hydrophobic in nature (Harris et al., 1991). Insoluble dietary fibres, evidenced by their insolubility in water or prepared physiological buffers, are largely hydrophobic as well, such as lignin and suberin. Because of the aqueous colonic environment, hydrophobic interactions between the highly hydrophobic carcinogenic compounds and hydrophobic dietary fibre components are expected and observed in vitro (Smith-Barbaro et al.., 1981; Harris et al., 1991). Further, non-lignified cell walls and dietary fibre components which are often degraded in the colon seemed to bind carcinogens less well, but still significantly (Harris et al., 1991). It is possible that the sequestering of hydrophobic carcinogens precludes their interactions with colonic mucosa. While it is not necessarily a given that bound carcinogen will have a protective effect by shielding colonic cells from them, studies have shown that wheat bran does inhibit the activation of cytochrome P450 systems, known to activate several potent carcinogens, such as 3-methyl-cholanthrene (Kawata et al., 1992), dimethylhydrazine (Smith-Barbaro et al., 1981b), and benzo[a]pyrene, a potent carcinogen commonly found in charred meat (Clinton and Visek, 1989).

Dietary fibres have been shown also to bind bile salts and bile acids (Monro et al., 1992). Conjugated bile acids may be protected from bacterial enzymes which would normally effect the conversion of primary bile acids to secondary acids. Bile acids which are conjugated to dietary fibres and passed through the alimentary canal unaffected therefore presents a mechanism by which secondary bile acids are prevented from interacting with colonic mucosa.

There are also mechanisms which have been proposed that indirectly involve effects of dietary fibres, such as altering the colonic environment, encouraging beneficial intestinal microflora, and by the actions of metabolites of dietary fibres expelled by these microflora. One such mechanism, proposed by Stephen and Cummings (1980) is by a raw increase in bacterial populations. Increases in bacterial populations, particularly beneficial microorganisms, have several potentially protective effects. A growth in bacterial populations implies an increase in the bacterial biomass, which itself increases stool mass significantly. Harris et al. (1999) point out that bacteria consist of, on average, roughly 80% water. This liquid displacement is responsible for the significant increase in stool mass from bacterial populations, which themselves represent relatively low dry mass. As outlined above, increases in stool mass are generally considered beneficial and protective. Since this involves an increase in available nutrients for these bacteria, it is specifically the degradable dietary fibres, such as those from fruits and vegetables, that increase stool mass by means of increasing bacterial populations. Bacterial populations also serve to lower the colonic environmental pH through fermentation waste products, such as short-chain fatty acids (Wijnands et al., 1999; MyIntyre et al., 1993). The solubility of free bile salts, known to exert their tumour-promotive effects on cells in which carcinogens have already caused DNA lesions, are known to have a solubility inversely proportional to pH–a lowering of colonic environmental pH will therefore reduce the opportunities of these bile precipitates to interact with these tumour-susceptible cells. A lowering of colonic pH may have other beneficial effects, however. Thornton (1981) cites that 7-a-dehydroxylase, a bacterial enzyme critical in the degradation of bile acids to secondary bile acids is inhibited below pH 6-6.5. Acidification of the colonic environment therefore not only causes precipitation of bile acids to a solid form, reducing contact with colonic mucosa, but also inhibits the detrimental degradation of the primary bile acids to secondary bile acids by intestinal microflora. A similar mechanistic approach was proposed by Newmark and Lupton (1990) involves the binding of calcium to free bile acids. At pH 6, calcium phosphate exhibits higher solubility than at pH 8. A physiological lowering of environmental pH could therefore increase the available calcium which can sequeseter free bile and fatty acids. Epidemiological studies have shown that populations with lower fecal pH exhibit overall lower rates of colon cancer, which supports these hypotheses. More direct interevention studies will elucidate the precise relationship or relationships between colonic pH and colon cancer. Butyrate, a specific short-chain fatty acid exhibits additional benefits by slowing the metabolic processes within colonocytes, and is the preferred energy source (Roediger, 1982). In vitro studies have shown that this relationship causes a decrease in the turnover rates of the already-high turnovers of colonocytes (Kim et al., 1980).

An in vivo study by Wijnands et al. (1999) confirmed that non-fermentable cellulose confers little protection against colorectal tumours, while fermentable polysaccharides which yield short-chain fatty acids confer a significant, positive protective effect. In contrast to the epidemiological studies (Hill, 1998; Gaard et al., 1996; Cummings et al., 1992; Burkitt, 1970), Wijnands et al. employed rats as a physiological model to approximate the efficacy of two dietary fibre components in preventing carcinogen-induced colonic tumours. Using Winstar rats (Charles River Wiga, Sulzfeld, Germany) , Wijnands et al. induced tumours with 1,2-dimethylhydrazine (DMH) and observed the effects of a high- and low-fat and -fibre diet, as well as the differential effects of fermentable versus non-fermentable fibre sources.

Using 468 eight-week old male Winstar rats, divided into 12 groups of 39 each, each of the groups were fed a diet of low or high cellulose (4-5% versus 22-24%), low or high galacto-oligosaccharides (8% versus 25%; a fermentable fibre source), and low (~3%), medium (7%), or high (15%) fat consisting of high-oleic sunflower oil. They found that the lowest fecal output was observed in the high-fat, high galacto-oligosaccharide-fed group. All other groups seemed to produce relatively similar nominal fecal masses. The highest fecal outputs were observed in the high cellulose diets. Because cellulose is a non-fermentable fibre, not degraded by endogenous enzymes, it often leaves the alimentary tract largely unaltered, resulting in increased fecal dry mass. Further, because it is not processed by endogenous metabolic enzymes, it contributes little to the energy input of the organism, and there is a greater stimulus to ingest more food, also resulting in a net increase in fecal output.

After treatment with 50mg/kg body weight DMH (once per week for 10 weeks; Sigma, Brussels, Belgium), colorectal tumours were observed with a mean incidence of 89%, indicating the apt choice of DMH as a carcinogen. It was found that the number of tumours increased significantly and consistently with an increase in fat content in the diet. In the case of the fermentable galacto-oligosaccharide diet, increases in fat content also yielded increases in tumour incidence, yet an increase in GOS did not yield statistically relevant decreases in tumour incidence. Significant, however, was the multiplicity of the tumours in each subject. While high cellulose-fed rats exhibited a relatively high number of tumours per individual, high GOS-fed rats exhibited a substantially decreased number, relative to their high-cellulose counterparts. The location of the tumours was found to be consistently in the distal two thirds of the colon, not influenced by diet variations. Wijnands et al. (1999) posited that short-chain fatty acid, a waste product of fermented galacto-oligosaccharides, not only was able to mediate the high cellular turnover of colonic cells, but also to provide an attractive energy source to the beneficial microflora, such as bifidobacteria and lactobacilli. This study is largely dependent upon a number of assumptions. The most fundamental assumption taken in this study is that rat digestive enzymes and intestinal microflora are similar those found in humans. Further, that the mechanisms by which the rats develop colorectal tumours, involving the K-ras gene in humans in most cases, is the same. Dimethylhydrazine was used to incite tumour development in the rats, assuming that the activity of subcutaneously-injected DMH accurately simulates the tumour-inducing properties of environmental and food-borne carcinogens, such as benzo[a]pyrene or 1,8-dinitropyrene. The use of isolated cellulose or GOS diets considers the dietary fibre components in isolation, which is obviously not the normal form of dietary fibre most humans ingest. Lignin, previously purported to be a significant cancer-protective dietary fibre component, was not considered in this study, presumably because the study considered only the rates of cancer incidence given environmental stimuli to incite tumour development. Indeed, subcutaneous administration of the carcinogen disregards many of the significant cancer-protective factors, such as hydrophobic interactions between carcinogens and fibre constituents. Many incidences of colorectal cancer in humans further depend on DNA lesions caused by carcinogens when administered physiologically through the alimentary canal. This study is therefore unable to consider the important effects of pH, carcinogen-sequestering, and carcinogen dilution by dietary fibre intakes, considered to be significant factors in the protective effects of dietary fibres against colon cancer. The use of high-oleic sunflower oil further makes the assumption that this oil adequately models, in general, the in vivo interactions of dietary fats. In fact, this disregards the oxidative effects of dietary lipids which can be significant (Latham et al., 1999) as well as differences in lipid metabolism between sunflower oil and many of the fats found in the typical Western diet, such as trans-fatty acids, high animal fat content, and popular use of highly unsaturated fatty acids, which are more prone to oxidation than more saturated fats. In fact, Latham et al. (1999) reported that dietary fish oils (n-3 polyunsaturdated fatty acids), such as eicosapentaneoic acid and docosahexaneoic acid were able to protect against carcinogenesis by dimethylhydrazine by increasing the apoptotic response in rat colonic cells. This study demonstrates that the dietary composition of fatty acids is an important factor in the consideration of colonic mutagenesis and supports the theory that the use of sunflower oil represents a gross over-simplification with respect to modeling a varied fat source diet with a single plant source oil. The relevance of the preventive effects of n-3 polyunsaturated fatty acids, such as fish oils, could be easily assayed by a cohort study involving populations where fish consumption, particularly consumption of fish high in n-3 polyunsaturated fatty acids, such as salmon, are compared in isolation to dietary fibre intake for colon cancer incidence. The mechanism by which highly unsaturated fatty acids, which are extremely prone to oxidation, is believed to protect against colon cancer by inducing apoptosis is through oxidative stress leading to loss of cellular viability. While sunflower oil is an unsaturated oil, it is unlikely that it is able to exert any cancer preventive effects by the same mechanism.

Dietary fibre, with its many potential colorectal cancer protective effects, has been shown to have cancer-promotive properties, as well. These cancer-promotive properties may be responsible for some of the equivocation between studies, some of which tout dietary fibre as a cancer-preventive measure, the supplementation of which to be advocated, some of which warn against fibre supplementation without proven just cause, such as in the case of constipation and psyllium supplementation (Wasan and Goodlad, 1996; Slavin and Ferguson, 1987). Although no studies have implicated dietary fibre in the direct increase of carcinogenicity of known DNA-damaging compounds, it has been postulated that dietary fibre, with bound carcinogen, could potentially be degraded by bacteria. While this requires that the carcinogens be present already from exogenous factors, such as charred meats or proteinaceous metabolic byproducts, this also presents an opportunity to concentrate these newly-released carcinogens within one location within the intestines due to the specific location of bacteria within the alimentary tract. One study (Harris et al., 1993) demonstrated that fibre-bound DNP (1,8-dinitropyrene), a known carcinogen, could be re-precipitated by degradation of the polysaccharides to which it had been bound. Insofar as bacteria within the colon are not homogeneously distributed, carcinogens could be concentrated and potentiated within a small region of the intestines, resulting in an increased probability of DNA damage. It is logical to believe that other bound carcinogens could potentially be concentrated within one area of the colon by the same mechanism. The same principle, when applied to bile acids, could enhance tumourgenicity in colonocytes which have already sustained DNA damage, also enhancing the development of colorectal cancer.

One perspective that most dietary fibre studies fail to address is the possibility that phenolic compounds, themselves a major component of lignin, contribute antioxidant properties to stool mass. It is well known that d-a-tocopherol, a specific phenolic, is a potent antioxidant due to free-radical quenching. Phenolics, in general, are good antioxidant compounds and are used even as high-temperature antioxidant additives in motor oils for the same reason. While most of the anti-cancer properties attributed to dietary fibre have been with respect to carcinogenic compounds, it is possible, with the high cellular proliferation and high metabolic rates of colonocytes, that metabolic free radicals, such as superoxide radical, are being produced at higher-than-normal rates. These radicals could be quenched by phenolic substances present in dietary fibres, such as lignin, or by other compounds present in dietary fibres or vegetal foods. Alabaster et al. (1996) pointed out that not only phytic acid, a significant constituent of many fibre-containing plants could significantly reduce colonic tumour incidences, but also that d-alpha-tocopherol, beta-carotene, and folic acid–non-fiber constituents of many fibre-containing foods–had significant protective effects. Many discrepancies and equivocations between studies may be attributed to neglecting non-dietary fibre compounds that are commonly found in fibre-rich diets. Indeed, this seemed to be the case of thiamine in the proposed effects of dietary fibre in the mediation of glucose metabolism. In fact, Alabaster et al. (1996) pointed out that beta-carotene is one of the most potent quenchers of singlet oxygen, and a widely used antioxidant added to lipid-containing foods to prevent oxidation. This is particularly relevant to the Western diet, characterized by low dietary fibre and high dietary fats. At these concentrations, dietary fibre-containing foods supplement with or high in beta carotene may present strong protective effects against colon cancer, but cautioned that further studies to elucidate possible mechanisms of protection by beta carotene are required before it can be attributed significant protective effects within the context of a high fibre diet.

It is highly likely that the observed cancer protective effects are due not to any one of these factors, in isolation, but rather to a combination of these factors, or a synergistic relationship between them.

SUPPLEMENTATION and DIET

With the present data that seem to indicate a positive preventive role of dietary fibre against colorectal cancer, it seems natural that the commercial food industry would exploit the opportunity to market food dietary fibre additives with the claim that they may play a role in cancer prevention. Indeed, cancer preventive studies cannot even ascertain which dietary fibre components–or even non-fibre components–are responsible for the observed trend towards lowering rates of colon cancer in a population. It is apparent from epidemiological studies (Caygill et al., 1998; Hill, 1998) that an increase in dietary fibre intake, from a wide variety of sources, coupled with a decrease in high-fat or high-protein foods leads to an overall decrease in statistical likelihood of developing colon cancer. There are also genetic factors that must be considered when determining if dietary fibre supplementation is adequate or appropriate for a given individual. Indeed, genetic factors, themselves not dependent on geography, tend to be grouped geographically simply because of the lack of positive random assortive matings within and between populations. The African diet studied by Burkitt (1970) could therefore involve regional hereditary and environmental factors that could offset, increase, or have no effect upon colon cancer rates with respect to dietary fibre intake. Wasasn and Goodlad (1996) point out, however, that it is common for an individual migrating from a low-risk population to a high-risk one to take on the risk of the new population, indicating that cultural, environmental, and dietary factors are more significant than hereditary ones. It follows naturally, therefore, that a dietary increase in fibre intake, as opposed to a supplement approach, cannot be contraindicated, and in fact, statistically lessens the chances of developing colorectal cancer.

Slavin and Ferguson (1987) outline the common compounds considered to belong to the Dietary Fibre family as follows:

Table 2 Common food sources of various fibre components

Cellulose

Hemicellulose

Pectin

whole-wheat flour

bran

cabbage family

peas and beans

apples

root vegetables

bran

cereals

whole grains

apples

citrus fruits

strawberries

Gums

Lignin

oatmeal

dried beans

legumes (soy, etc.)

mature vegetables

wheat

Because of the discrepancies between dietary fibre content reporting methodologies, it is difficult to ascertain the precise dietary fibre content in foods or to establish a precise dietary fibre intake requirement. Legumes, for example, are known for a high dietary fibre content, but Slavin (1987) points out that much of this reported dietary fibre may, in fact, be residual protein or starch components, due to their tremendously high protein and starch content. It is only natural that a certain portion of this protein or starch content be undegradable both physiologically and in assays. For stool bulk purposes, it is not relevant, because undegraded proteins and starches seem not to contribute to cancer development but neither can it be claimed to have preventive effects, other than dilution by stool bulk. It is unfortunate, however, due to Western cultural dietary practices as well as Western food processing practices, that it is extremely difficult, even for a vegetarian diet, to obtain high levels of fibre and still maintain a balanced diet. The issue may be more complex, however, than simply boosting fibre levels. Dietary fibre is known to interfere with absorption of more than just carcinogens–it can interfere with nutrient absorption in the case of vitamins or minerals. Increasing fibre intake as part of an overall dietary modification must be done responsibly. Because of the ability of dietary fibres to interfere with nutrient absorption, it might be advisable to coordinate calcium and other nutrient intakes with a time that does not coincide with significant dietary fibre intakes. Conversely, it seems advisable to coordinate dietary fibre intake with intake of foods high in dietary fats or proteins, that the fibres would be present in the colon to interact with potential carcinogens. It would be most prudent, of course, to distribute fibre intake throughout the day, with the greatest concentration in conjunction with high-fat or high-protein foods. This is not always possible, however, due to the nature of most foods high in dietary fibre. Cultural preferences in the Western diet seem to place great emphasis on high-fat, highly-processed, high-protein foods. Further, barbequed meats, a common item on the menu of many Western diets, provides still more carcinogens from the charring of meats, such as benzo[a]pyrene. Indeed, one of the dangers of Western cultures is to seek a “cure-all”–a single treatment that encourages hedonism and an irresponsible approach to dietary planning. It must be stressed that dietary fibre, in spite of its apparent cancer protective effects, cannot act as a crutch, cannot replace responsible dietary planning. The marketing of fibre-supplemented foods without adequate research into the precise mechanisms thought to be involved in cancer prevention–further the marketing of supplemented foodstuffs without the education to promote a balanced diet wherein dietary fibre fills a specific niche–can only be considered detrimental to the health of the population, as a whole. It is important that increases in dietary fibre intake be effected gradually. Dietary fibre changes the colonic environment drastically with respect to pH, stool bulk, transit time, and microflora. A sudden change in dietary fibre intake can cause severe gastrointestinal distress in the individual, as well as a drastic upset in the intestinal microorganismal environment.

Fibre supplementation has long been the common remedy for constipation. And it is indicated for such cases. But non-therapeutic supplementation, such as food products marketed as “fibre-enriched” may be misleading to the consumer, and may, in fact, be detrimental. As previously determined, only specific types of fibres are effective in cancer prevention. Further, Wasan and Goodlad (1996) point out that the artificial and unnecessary increase in carbohydrates can stimulate the already-high colonic epithelial cell proliferation rates. Short-chain fatty acids, a metabolic waste product of endogenous colonic bacteria are effective at inducing cellular apoptosis in pre-cancerous cells (Harris and Ferguson, 1999; Wasan and Goodlad, 1996).

Dietary fibre has long been known to be beneficial to health. Since classical Greek civilizations, it has been used therapeutically to relieve constipation. More recently, there has been much supporting evidence that certain specific dietary fibres are effective in the prevention of colorectal and breast cancers. Given the research into dietary fibre as a preventive factor against colon cancer, it is prudent to advise a conscious increase in dietary fibre intake from a variety of sources, a lowering of dietary fat intakes from all sources, and coordination of dietary fibre intake with that of other nutrients to optimize dietary fibre-nutrient interactions. It is apparent from current research that the interactions in which dietary fibre is involved to effect this apparent protective effect are complex and operate on different cellular systems spatially, temporally, and physico-chemically. Indeed, dietary fibre comprises a heterogeneous mixture of compounds, some of which may exert wholly-different effects upon colonocytes, intestinal microflora, colonic environment, carcinogens, and tumour-inducers. Further research is necessary into the precise mechanisms involved in the observed cancer preventive effects to elucidate present hypotheses. Epidemiological studies have been completed, the majority of which indicate a positive protective effect of dietary fibre on colon cancer. Studies involving the isolated compounds of dietary fibres should first be conducted to ascertain the nature of the cellular interactions involved in the development and suppression of tumours. Cohort studies with monitored dietary intakes should then provide more insight into physiological effects of dietary fibre, having gained insights into the mechanistic factors underlying the observed results of the cohort studies.

References

Alabaster O, Tang Z, and Shivapurkar N. 1996. Mut Res. 350:185.

Anderson JW. 1992. Diabetes Research and Clinical Practice. 17:71

Bakker SJL, Hoogeven EK, Nijpels G, Kostense PJ, Dekker JM, Gans ROB, and Heine RJ. 1998 Diabetologia. 41:1168.

Brown L, Rosner B, Willett WW, and Sacks FM. 1999. Am J Clin Nutr. 69:30.

Burkitt DP. 1970. Lancet. 2: 1237.

Caygill CPJ, Charlett A, and Hill MJ. 1998. Eur J Cancer Prev. 7:S11.

Clinton SK and Visek WJ. 1989. J Nutr. 119:395.

Cummings JH, Bingham SAA, Heaton KW, and Eastwood MA. 1992. Gastroenterol. 103:1783.

Gaard M, Tretli S, and Loken EB. 1996. Eur J Cancer Prev. 5:445.

Graf E and Eaton JW. 1993. Nutr. Cancer. 19:11.

Ferguson LR and Harris PJ. Cancer Prev. 8:17.

Furda I. 1977. Cereal Foods World. 22:252.

Harris PJ, Ferguson RL. 1999. Mutation Res. 443:95

Harris PJ, Roberton AM, Hollands HJ, and Ferguson LR. 1991. Mutation Research. 260:203.

Heckman MM and Lane S. J Assoc Off Anal Chem. 64:1339.

Hellendoorn EW, Nordhoff MG, Stagman J. 1975. J Sci Food Agric. 26:1461.

Hill MJ. 1998. Eur J Cancer Prev. 7:S5

Kawata S, Tamura S, Matsuda Y, and Ito N. 1992. Carcinogenesis. 13:2120

Kritchevsky D. 1998. Eur J Cancer Prev. 7:S-33.

Jarvis MC. 1984. Plant Cell Environ. 7:153.

Kim YS,Tsao D, Siddiqui B, Whitehead JS, Bennett J, and Hicks J. 1980. Cancer: 45.

McIntyre A, Gibson PR, and Young GP. 1993. Gut. 34:386.

Monro JA, Lee J, and Sinclair BR. 1992. Food Chem. 44:325.

Newmark HL and Lupton JR. 1990. Nutrition and Cancer. 14:161.

Ross JK, English C, Perlmutter CA. J am Diet Assoc. 85:1111.

Schneeman BO. 1993. Am J Clin Nutr. 69:1.

Schneeman BO. 1986. Food Technol. 40:104.

Schneeman BO. 1999. Am J Clin Nutr. 69:1.

Sills RC, Hong HL, Melnick RL, Boorman GA, and Devereux TR. 1999. Carcinogenesis. 20:657.

Slavin JL. 1987. J Am Diet Assoc. 87(9)

Smith-Barbaro PD, Hansen D, and Reddy BS. 1981. J. National Cancer Inst. 67:495.

Smith-Barbaro PD, Hansen D, and Reddy BS. 1981b. 67:495.

Southgate DAT. 1969. J Sci Food Agric 26:1461

Stephen AM and Cummings JH. 1979. Nature. 284:283.

Thornton JR. 1981. Lancet. 1: 1081.

Van Soest PJ. 1978. Dietary Fibres: their definition and nutritional properties. Am J Clin Nutr.31:512.

Wijnands MVW, Appel MJ, Hollanders VMH, and Woutersen RA. 1999. Carcinogenesis. 20:651.