Dietary fibre and health: an overview

Transcription

1 REVIEW Dietary fibre and health: an overview J. L. Buttriss and C. S. Stokes British Nutrition Foundation, London, UK Summary Many studies have found that people on diets high in fibre have reduced risks of certain diseases such as cancers, coronary heart disease, obesity and possibly diabetes. Fibre is a collective term for a group of compounds, which differ in their chemical structure and physical properties and elicit a variety of physiological effects. Some health benefits linked to fibre consumption are well established (e.g. promoting a regular bowel habit) and others are becoming more firmly established. The effects of the various non-digestible components are not always fully interchangeable, although there is considerable overlap, but they share an ability to pass undigested into the large bowel. This makes it important that fibre intake comes from a range of sources to ensure maximum health benefits. There are disagreements as to which plant-derived compounds constitute fibre and which analytical methodologies should be used to determine the fibre content of foods. This lack of a universal definition for these carbohydrates that resist digestion in the small intestine has led to complications in establishing and communicating consistent recommendations for the type and amounts of fibre components in the diet, as well as issues related to food labelling and health claims. Furthermore, current analytical methods may typically underestimate the fibre content of foods, using the definition advised by the European Food Safety Authority. Chemical nomenclature is necessary for a coherent and enforceable approach to the measurement of fibre (and related to this, product labelling), but reconciling chemical definitions for fibre with definitions that reflect physiological effects remains a major challenge. Progress in the application of the emerging evidence regarding dietary fibre and health is constrained by a need to resolve the definition and methodology issues. Most consumers are likely to be unaware of the more recent developments and the overlapping properties of different types of carbohydrate that resist digestion in the small intestine, and the particular sources of these carbohydrates. Long overdue is the translation of the growing knowledge about the physiological properties of carbohydrates, particularly non-digestible carbohydrates, into clear and unambiguous public health messages. Many terms in common use stem from chemistry rather than physiology, and this makes derivation of health messages more difficult. Correspondence: Professor Judith L. Buttriss, Director General, British Nutrition Foundation, High Holborn House, High Holborn, London, WC1V 6RQ, UK. j.buttriss@nutrition.org.uk 186

2 Dietary fibre and health 187 Meanwhile in the UK, where the main sources of fibre are cereals and cereal products, followed by vegetables and potatoes, adults are not meeting the recommended 18 g of fibre per day (defined in the UK as non-starch polysaccharide measured by the Englyst method). The recommended amount for the UK, established in 1991, is lower than elsewhere in the world and corresponds to about 24 g according to the method used by the Association of Official Analytical Chemists (the AOAC method). Reference values in other parts of the world are typically between 25 and 35 g (AOAC fibre). A diet rich in whole grains, pulses, fruit and vegetables will assist in meeting current fibre recommendations, but a growing number of other foods now contain carbohydrate-derived ingredients that resist digestion. Keywords: dietary fibre, non-starch polysaccharides, oligosaccharides, resistant starch Introduction Currently, there is no single definition for dietary fibre that is accepted worldwide. This is largely because of the disagreement about precisely which plant-derived substances should be included in the definition and, directly linked with this, the analytical methodology to be used to derive fibre values. This dilemma has hampered progress in the fibre research field and also, potentially, hampered the implementation and communication of research findings that could benefit health. Dietary fibre was first defined by Hipsley in 1953 as non-digestible components of plants that make up the plant cell wall, for example, cellulose, hemicelluloses [both non-starch polysaccharides (NSP)] and lignin. Another major development took place in 1976 when Trowell et al. recognised that other non-digestible polysaccharides (i.e. gums, mucilages) associated with the plant cell wall but not strictly a part of it also contributed to the dietary fibre content of the diet (Trowell et al. 1976). Hence, the definition was expanded. For some time, fibre was classified as either soluble or insoluble fibre, depending on whether it resisted fermentation in the large bowel (insoluble) or not (soluble) (Stephen & Cummings 1980). However, the soluble/insoluble categorisation has since been recognised to be misleading because some insoluble fibres are in fact fermented in the large bowel and solubility in water does not always predict physiological effects. Since the 1970s, interest has developed in other carbohydrate-derived components with physiological properties, in particular resistant starches and oligosaccharides. Indeed, with regard to resistant starch, the observation that the rate and extent of starch digestion can vary has been one of the most important developments in the understanding of carbohydrates in the past 30 years according to Cummings and Stephen (2007). Yet, neither of these classes of carbohydrate (that both resist digestion in the small intestine) fit completely into either category (soluble or insoluble). Over time, further refinements have been made to the published definitions, related both to physiological considerations (e.g. digestion, absorption, effects on health) and to methods of analysis. In recent years, various expert committees have grappled with the complexities of defining precisely what constitutes dietary fibre and exactly how it should be measured. Some definitions are restricted to NSP as measured by the Englyst method (Englyst & Cummings 1988; Englyst et al. 1996); this is the approach used to define the dietary reference value for fibre published in the UK back in 1991 (Department of Health 1991). NSP is defined as comprising cellulose, hemicelluloses, pectin, arabinoxylans, betaglucan, glucomannans, plant gums and mucilages and hydrocolloids, all of which are principally found in the plant cell wall (Cummings & Stephen 2007). However, other definitions additionally include oligosaccharides, resistant starch and resistant maltodextrins, all of which pass undigested through the small intestine into the colon. Table 1 provides further information on the fractions of non-digestible carbohydrates captured to varying degrees by existing definitions. Currently, several different official definitions for fibre are used around the world. A common feature of most, if not all, definitions is non-digestibility in the small intestine, that is, a situation whereby some carbohydrate

3 188 J. L. Buttriss and C. S. Stokes Table 1 Description and sources of the main dietary fibre fractions captured by current definitions Fibre component Description Food sources Cellulose Hemicelluloses Pectins Beta-glucans Resistant starch Non-digestible oligosaccharides (NDOs) Other synthetic carbohydrate compounds Gums and mucilages Lignin Other minor components Polysaccharides comprising up to closely packed glucose units, arranged linearly, making cellulose very insoluble and resistant to digestion by human enzymes. Polysaccharides containing sugars other than glucose. Associated with cellulose in cell walls and present in both water soluble and insoluble forms. Polysaccharides comprising galacturonic acid and a variety of sugars; soluble in hot water and forms gels on cooling. Glucose polymers that, unlike cellulose, have a branched structure enabling them to form viscous solutions. Starch and starch degradation products that are not absorbed in the small intestine. Four classes have been identified: physically inaccessible starch, RS1; native starch granules, RS2; retrograded starch, RS3; and chemically modified starch, RS4 (Englyst & Cummings 1987). NDOs comprising three to ten sugar units occur naturally in plants consumed as foods, mainly vegetables, cereals and nuts. Can also be made chemically or enzymatically from mono- or disaccharides or by enzyme hydrolysis of polysaccharides. Because they are non-digested, they exhibit similar physical effects to their larger polysaccharide counterparts. They are typically fermentable and some have prebiotic properties [e.g. fructans such as FOS obtained from inulins with a degree of polymerisation of 3 60 and synthetic analogues synthesised from sucrose]. Physiological properties have been confirmed for some NDOs and these are primarily mediated via change to the gut microflora (either composition or activity). Synthetic derivatives of cellulose (for example, methyl cellulose and hydroxypropylmethyl cellulose) are non-digestible and, unlike their parent (cellulose), are soluble. But they are hardly fermented by microflora. Polydextrose has an average degree of polymerisation of 12 and is synthesised from glucose and sorbitol. It is partially fermented in the colon (~50% in humans) and has bulking and prebiotic properties. Gums are hydrocolloids derived from plant exudates. Mucilages are present in the cells of the outer layers of seeds of the plantain family, for example pysillium. Both are used as gelling agents, thickeners, stabilisers and emulsifying agents. Not a polysaccharide but chemically bound to hemicelluloses in plant cell walls. Phytic acid (inositol hexaphosphate) is associated with fibre in some foods, especially cereal grains. May reduce mineral absorption in the small intestine as it binds strongly. Other compounds associated with fibre include tannins, cutins and phytosterols. Principal component of the cell walls of most plants. Forms about 25% of the fibre in grains and fruit and about a third in vegetables and nuts. Much of the fibre in cereal bran is cellulose. Forms about a third of the fibre in vegetables, fruits, legumes and nuts. The main dietary sources are cereal grains. Found in cell walls and intracellular tissue of fruits and vegetables. Fruits contain the most, but pectins also represent 15 20% of the fibre in vegetables, legumes and nuts. Sugar beet and potatoes are sources. Major component of cell wall material in oats and barley, only present in small amounts in wheat. Legumes are one of the main sources of RS1 (because of their thick cell walls). Unripe bananas provide RS2 as do high amylose starches (produced industrially). RS3 is produced during cooking, cooling and storage of foods (e.g. potatoes). Exact quantification is difficult because of the impact of cooking and storage. Onions, chicory and Jerusalem artichokes are the major dietary sources. Currently, use in food of FOS and galacto-oligosaccharides is permitted in most European countries. (There has been recent discussion about the degree of polymerisation considered necessary to justify inclusion in the fibre definition. In the current proposal from the European Commission, only saccharides with three or more units are included, there being a concern about laxative effects with smaller molecules.) Polydextrose, for example, is used in some reduced energy products as a bulking agent to replace sugars in foods and to provide texture. Its contribution to energy is lower at just 1 kcal/g (Auerbach et al. 2007). Gums: plant exudates (gum arabic and tragacanth), seeds (guar and locust beans) and seaweed extracts (agar, carageenans, alginates). Mucilages: for example, pysillium. Foods with a woody component, for example celery, and the outer layers of cereal grains. Cereal grains. Source: Derived from text in Gray NDOs, non-digestible oligosaccharides; FOS, fructo-oligosaccharides.

4 Dietary fibre and health 189 passes undigested from the ileum into the colon. The majority are in accord with methods of analysis approved by the Association of Official Analytical Chemists (AOAC), for example the enzymatic gravimetric method (method ), although definitions differ in relation to minor constituents. For example, the American Association of Cereal Chemists (AACC 2001) uses the following definition, which includes oligosaccharides: Dietary fibre is the remnants of the edible part of plants or analogous carbohydrates that are resistant to digestion and absorption in the human small intestine, with complete or partial fermentation in the large intestine. Dietary fibre includes polysaccharides, oligosaccharides, lignin and associated plant substances. Dietary fibres promote beneficial physiological effects including laxation, and/or blood cholesterol attenuation, and/or blood glucose attenuation. In its 2001 definition, the Food and Nutrition Board of the Institute of Medicine in the USA has separated traditional NSP fibre (that they refer to as dietary fibre ) from the other compounds that have been attracting considerable research attention in recent years, which they described as functional fibre. They stated that dietary fibre consists of non-digestible carbohydrates and lignin that are intrinsic and intact in plants and that functional fibre consists of isolated, nondigestible carbohydrates that have beneficial physiological effects in humans. Total fibre is the sum of dietary fibre and functional fibre (FNB 2001). For more than 15 years, the international CODEX Alimentarius Commission has debated a definition that can be used universally. Through the eight-stage CODEX process, a definition has been developed and consulted upon that describes dietary fibre as carbohydrate polymers with a degree of polymerisation not lower than three (to exclude mono- and disaccharides) and includes (1) naturally occurring edible carbohydrate polymers; (2) carbohydrate polymers obtained from food raw material by physical, enzymatic or chemical means; and (3) synthetic carbohydrate polymers. This definition reached Step 7 of the CODEX process, reflecting a consensus opinion, and has gathered wide support around the world but is noteworthy in that it is not consistent with the NSP-based definition adopted by the UK Department of Health in 1991, which includes only plant cell wall-associated fractions. Another CODEX meeting is scheduled for November Ultimately, it will be the European Commission (EC), rather than CODEX, that will determine the rules for the European Union (EU) and hence Britain. In July 2007, the European Food Safety Authority (EFSA) provided an opinion to the Commission on this issue (EFSA 2007), which notes that the main problem in making a differentiation between NSP fibre and functional fibre is that no analytical method differentiates between them once they are mixed in a food product. Similarly, NSP from plant cell walls cannot be distinguished from added NSP with a similar chemical structure (EFSA 2007). The EFSA Panel recommends that dietary fibre should include all non-digestible carbohydrates because of the key importance of digestibility in the small intestine for the nutritional effects of carbohydrates in humans. EFSA defines dietary fibre as non-digestible carbohydrates plus lignin that comprises: NSP cellulose, hemicelluloses, pectins, hydrocolloids (gums, mucilages, beta-glucans); resistant oligosaccharides fructo-oligosaccharides (FOS), galacto-oligosaccharides (GOS), other oligosaccharides that resist digestion (with three or more monomeric units); resistant starch physically enclosed starch, some types of raw starch granules, retrograded amylose, chemically and/or physically modified starches; lignin naturally associated with dietary fibre polysaccharides. The EFSA opinion also considered methods of analysis, noting that several methods need to be applied in tandem in order to capture all of the above fractions because none of the methods currently available are optimal for measuring the range of components now generally accepted as contributing to dietary fibre. For example, resistant oligosaccharides and inulin are not captured by any of the current methods for total dietary fibre and have to be measured separately. Therefore, the current methods almost certainly underestimate the fibre content of foods, using the definitions of dietary fibre advised by EFSA and CODEX. The development of a new integrated method to measure all fibres as advised by EFSA and CODEX is currently underway (McCleary 2007). AOAC interlaboratory investigation of the method started in March 2008 and is planned to be completed by the end of This method would allow measurement of all fibre types with a single AOAC method, avoiding possible overlap of analysis that might potentially occur with multiple methods. In line with EFSA s opinion and also the definition proposed by CODEX, in January 2008, the EC published a working draft amendment to Directive 90/496/ EEC, which defines dietary fibre. The Commission acknowledges that fibre has traditionally been consumed as plant material, with associated health effects, but recent evidence has shown that similar beneficial physiological effects may be obtained from other carbo-

5 190 J. L. Buttriss and C. S. Stokes hydrate polymers that are not digestible and do not occur naturally in food (European Commission 2008). The Commission defines fibre as carbohydrate polymers with three or more monomeric units, which are neither digested nor absorbed in the small intestine. Dietary fibre consists of one or more of: edible carbohydrate polymers naturally occurring in the food as consumed; carbohydrate polymers that have been obtained from food raw material by physical, enzymatic or chemical means; synthetic carbohydrate polymers. With the exception of non-digestible edible carbohydrate polymers that occur naturally in foods, the amendment also states that there should be evidence of a beneficial physiological effect in the generally accepted scientific evidence to support the inclusion of other material captured by the definition. The document also advocates a value of 2 kcal/g (8 kj/g) for the energy value of dietary fibre. At the time of writing, this definition has largely passed through the consultation phase. It has passed through the Standing Committee with one abstention and is to be presented to the European Parliament. Central to discussions about the definition of dietary fibre is the extent to which emerging evidence supports a role for non-nsp fibre in the health effects associated with total dietary fibre. Back in 2005, the Food and Agriculture Organisation/World Health Organization (FAO/WHO) agreed to undertake a scientific update on carbohydrates in human nutrition, the findings of which were published at the end of 2007 in a supplement to the European Journal of Clinical Nutrition (Nishida et al. 2007). In contrast to the recommendations of CODEX and EFSA (and a number of other scientific regulatory, analytical and trade bodies), the WHO/FAO working group specified that the term dietary fibre should be reserved for cell wall polysaccharides of vegetables, fruits and whole grains, the health benefits of which have been clearly established, rather than synthetic, isolated or purified oligosaccharides with diverse, and in some cases, unique, physiological effects. So, currently, there are two distinct schools of thought on definitions and methodology, which boils down to: NSP vs. a wider definition of fibre as measured by the AOAC method. On the one hand, some believe that dietary fibre labelling should provide guidance about largely unrefined plant foods shown to be associated with bowel health. Whereas, others accept that those non-nsp types of carbohydrate that resist digestion in the small intestine also act as substrates for short-chain fatty acid (SCFA) production and are capable of contributing to some of the physiological effects associated with highfibre intakes that are described later and may have unique properties of their own, and can therefore make a contribution to the impact of the fibre content of the diet. In the absence of a final decision on the definition and the associated analytical methods, confusion continues. This serves to undermine the scientific evidence supporting an important role for dietary fibre in health and disease prevention, and carries with it the likelihood that some consumers may be misled as a result of the ambiguities that the confusion generates. If the Englyst method is used to derive food table values (as has been the case in the UK), these published values may not equate with those that appear on food labels, calculated using the AOAC method (the method used globally and advocated by the EC to ensure consistent labelling of products). This is particularly the case for foods rich in starch, such as potato, bread, beans and breakfast cereals such as cornflakes, for which the AOAC-derived value can be considerably greater than the NSP value (Cummings & Stephen 2007). At best, this is confusing and it makes the comparison of food labelling information with dietary targets for individuals or population groups difficult, as account has to be taken of which method has been applied in generating the fibre value, and this is not necessarily obvious at first glance. The lack of a decision about the definition of fibre also makes it difficult for the food industry to interpret existing EU law (e.g. the Regulation concerning nutrition and health claims; European Commission 2007) for products sold in the UK, as there is lack of clarity on the analytical method. Dietary fibre in the UK UK recommendations The recommended dietary fibre intake in the UK is set at 18 g per day (Department of Health 1991). This was derived using the fraction of non-digestible material described as NSP that is measured by the Englyst method and is consistent with the recent FAO/WHO opinion but lower than recommendations elsewhere in the world. Fibre recommendations vary considerably around the world, reflecting differences in the way that dietary reference values are defined but also the considerations about definition and analytical methods referred to earlier. Even with these limitations, it is evident that the UK recommendation is among the lowest in the world. Reference values in other parts of the world are typically expressed as AOAC fibre and are

6 Dietary fibre and health 191 Table 2 Sources and intake levels of dietary fibre (measured as non-starch polysaccharide) in the UK Sources of fibre, % of total intake Children 4 18 years* Adults years Adults 65+ years Boys Girls Men Women Men Women Cereals and cereal products Vegetables Potatoes Savoury snacks Fruit and nuts Average daily intake 11.2 g 9.7 g 15.2 g 12.6 g 13.5 g 11 g Source: *Gregory and Lowe (2000). Henderson et al. (2003). Finch et al. (1998). usually in the range of g, but in some cases range up to 40 g/day (see Lunn & Buttriss 2007). The figure of 18 g measured as NSP roughly equates with 24 g as measured by the AOAC methodology, and the average intake of adults in the UK is currently of the order of 13 g/day. The UK government s Scientific Advisory Committee on Nutrition (SACN) has established a working group to consider dietary carbohydrates and health. The working group s remit includes reviewing the role of dietary carbohydrate in colorectal health (including colorectal cancer, irritable bowel syndrome and constipation), dietary carbohydrate and metabolic health [including insulin resistance, glycaemic index (GI) and obesity], the evidence for dental health which has arisen since the COMA report (Department of Health 1991), and the terminology, classification and definitions of types of carbohydrate in the diet. Additionally, to facilitate discussions around fibre, a major review is underway by experts in the carbohydrates field and is due to be completed this summer. A summary was published in June 2008 ( Sources of dietary fibre and intake levels in the UK In the UK, the main source of dietary fibre (measured as NSP) is cereal and cereal products, as illustrated in Table 2. The table also provides average intakes for different age groups using data from the National Diet and Nutrition Survey (NDNS) programme. Adults are not meeting the recommended 18 g of fibre per day, with men (aged years) consuming an average of 15.2 g of NSP fibre per day and women (aged years) 12.6 g. The NDNS also provides information on the range of NSP fibre intakes in the UK. In men, the range is 8 38 g/ Fibre (g/day) A B Nl Ger F I Gre S Figure 1 Fibre intakes: comparison between European countries. The grey bars represent the range of intakes of fibre (g/day) as determined by the mean of highest to mean of lowest quartiles/tertiles, in relation to the EURODIET recommendation (black bar). Source: Data taken from Gibney (2000). A, Austria; B, Belgium; NL, Netherlands; Ger, Germany; F, Finland; I, Ireland; GRE, Greece; S, Spain. day and in women, it is 5 24 g/day, expressed as the upper and lower 2.5 percentiles (Henderson et al. 2003). Figure 1 displays the range of intakes of AOAC fibre (g per day) for a number of EU countries. For comparison, the EURODIET project recommended g total fibre per day for European populations (Gibney 2000). Very little information exists about intakes of subfractions of non-digestible carbohydrates (Elia & Cummings 2007). In European countries, between 11 and 33 g/day of NSP is thought to reach the colon; higher intakes may occur in vegetarians. Resistant starch intakes are much more difficult to assess. Estimates for European populations range from 3 10 g per day but are very diet dependent; for example, an unripe banana will provide

7 192 J. L. Buttriss and C. S. Stokes Table 3 Fibre content of some commonly consumed grains Low (less than 3 g per 100 g)* Medium (3 6 g per 100 g)* High (6 g or more per 100 g)* Source NSP value (g/100 g) Source NSP value (g/100 g) Source NSP value (g/100 g) White rice 0.1 Wheat flour, white 3.1 All bran 24.5 Brown rice 0.8 Granary bread 4.3 Crispbread 11.7 Porridge 0.8 Puffed wheat 5.6 Oat bran flakes 10 Rice Krispies 0.7 Rye bread 4.4 Shredded wheat 9.8 Spaghetti, white 1.2 Spaghetti, wholemeal 3.5 Weetabix 9.7 White bread 1.5 Brown bread 3.5 Wheat flour, wholemeal 9 Source of composition data: FSA (2002). *For presentation purposes, the bands are based on the criteria contained within the European Commission Regulation on Nutrition and Health Claims (European Commission 2007).The analytical method associated with these values (3 and 6 g) has yet to be clarified in legislation, as indicated earlier. Currently, the UK food tables provide data in terms of NSP. NSP, non-starch polysaccharides. Table 4 Fibre content of fruits, vegetables and pulses Low (less than 3 g per 100 g)* Medium (3 g per 100 g)* High (6 g or more per 100 g)* Source NSP value (g/100 g) Source NSP value (g/100 g) Source NSP value (g/100 g) Broccoli 2.3 Baked beans 3.8 Red kidney beans 6.2 Carrots 2.5 Brussels sprouts 3.1 Figs 6.9 Apples (no skin) 1.6 Butter beans 5.2 Apricots (ready to eat) 6.3 Apples (with skin) 1.8 Chickpeas 4.3 Pears (no skin) 1.7 Lentils 3.8 Pears (with skin) 2.2 Avocado 3.4 Baked potato (no skin) 1.4 Passion fruit 3.3 Baked potato (with skin) 2.7 Source of composition data: FSA (2002). *For presentation purposes, the bands are based on the criteria contained within the European Commission Regulation on Nutrition and Health Claims (European Commission 2007).The analytical method associated with these values (3 and 6 g) has yet to be clarified in legislation, as indicated earlier. Currently, the UK food tables provide data in terms of NSP. NSP, non-starch polysaccharides. 10 g of resistant starch by itself. There are almost no data on intakes of non-digestible oligosaccharides; estimates suggest 1 10 g/day in European populations, although levels will again be very diet dependent. For example, 50 g of Jerusalem artichoke will provide 8 10 g. Polyols, for example sorbitol and xylitol, are used in energyreduced foods, and people selecting sugar-free products may easily consume 20 g/day (Elia & Cummings 2007). Hence, Elia and Cummings estimated that g of carbohydrate is likely to reach the colon in European populations, and this may rise to 50 g in population groups consuming large amounts of fruits, vegetables and whole grain cereals. Fibre content of foods As was evident from Table 2, fibre is found in foods such as grains, pulses, fruits and vegetables. The fibre content of foods derived from grains varies depending on the amount naturally present and also the degree of milling and processing. Whole grain foods are typically rich sources of dietary fibre. Table 3 summarises the fibre content of some commonly consumed grainderived foods, and Table 4 summarises the fibre content of some pulses, fruits and vegetables. In both cases, the foods are grouped using the thresholds contained within the EC Regulation on nutrition and health claims, whereby 3 g or more enables a source of claim and 6 g or more a high in claim (European Commission 2007). Fibre in health and disease People who eat fibre-rich diets tend to have a reduced risk of certain cancers, coronary heart disease (CHD) and obesity. Fibre can help to reduce the energy density of foods owing to its bulking effect and can promote

8 Dietary fibre and health 193 Table 5 Principal physiological properties of non-digestible carbohydrate fractions Non-digestible carbohydrate fraction Non-starch polysaccharides Non-digestible (non-alpha-glucan) oligosaccharides Resistant starches Principal physiological properties Increase stool output, increase satiety, cholesterol lowering (some forms only, highly viscous fibres such as beta-glucan and pectins) and source of SCFA (acetate, propionate and butyrate) Source of SCFA, alter microflora balance (i.e. act as prebiotics), immunomodulatory role (reported improvements in gut barrier function against infection), increase calcium absorption and possible role for prebiotics in cancer protection Source of SCFA, increase stool output, suggestion of a beneficial effect on glucose handing and possibly blood lipids, and some suggestion of a prebiotic effect Source: Based on data in Nugent (2005), Cummings and Stephen (2007) and Elia and Cummings (2007). SCFA, short-chain fatty acids. satiety, as described in more detail later. Although whole grain foods provide fibre, in addition, they supply micronutrients, essential fatty acids and some protein, all of which contribute to good health. They also contain a wide range of other bioactive substances that may in due course be shown to have important health benefits. Over the past 50 years or so, three key physiological characteristics have been demonstrated for diets rich in dietary fibre: prevention of constipation, lowering of blood cholesterol and effects on satiety. However, not all of the carbohydrate fractions classed as dietary fibre are able to perform each of these or to the same extent, as summarised in Table 5. Digestive health The principal carbohydrates that reach the human large intestine are NSP, resistant starch, non-digestible oligosaccharides and some polyols and modified starches (Elia & Cummings 2007). Also, in non-caucasian populations, who are typically lactase deficient, the disaccharide lactose can reach the large intestine (Buttriss & Korpela 2002). Fibre components are partially or completely fermented (i.e. used as an energy source) by the gut microflora in the large bowel (colon), the carbohydrate being converted into gases (hydrogen, methane and carbon dioxide) and SCFA (mainly acetate, propionate and butyrate). The availability of carbohydrate substrate in the colon results in an increase in the number of bacteria and hence an increase in faecal mass. Some fibre components absorb fluid and so increase stool weight, and this, together with the increase in microbial biomass associated with fermentation, largely determines stool weight. Cummings and colleagues have calculated the increase in stool weight attributed to different fibre sources and it is evident from the following data (expressed as the average amount of faecal mass for every gram of fibre consumed) that some forms are more efficient than others: raw bran 7.2 g/g, fruit and vegetables 6.0 g/g, psyllium 4 g/g, oats 3.4 g/g, legumes 1.5 g/g and pectin 1.3 g/g (Elia & Cummings 2007). The SCFA produced as a result of the fermentation of fibre are thought to be important for a number of reasons. Butyrate is the preferred fuel of the colonocytes that line the colon and may be a primary protective factor for the health of these cells and hence the colon. SCFA also lower the ph in the colon and thereby inhibit the growth of pathogenic organisms and also the formation of toxic breakdown products (see Gray 2006 and Scott et al. 2008). SCFA absorbed into the systemic blood supply constitute an energy source, as indicated earlier, but also provoke beneficial effects on lipid (and possibly glucose) metabolism. Most resistant starches and non-digested oligosaccharides are thought to be fermentable and some nondigested oligosaccharides, for example inulin and FOS, exhibit prebiotic properties, that is they specifically stimulate the growth in numbers of the beneficial gut bacteria, bifidobacteria and lactic acid bacteria (Gibson et al. 2004; see Nugent 2005). There is also limited evidence for a prebiotic effect of some starches that resist digestion. Prebiotics A prebiotic is defined as a non-digestible food ingredient that beneficially affects the host by selectively stimulating the growth and/or activity of one or a limited number of bacteria in the colon, and thus improves host health (Gibson & Roberfroid 1995). Compared with other carbohydrates that resist digestion, prebiotics can be distinguished by their particular fermentation pattern

9 194 J. L. Buttriss and C. S. Stokes and selective stimulation of the growth of colonic bifidobacteria that are capable of producing butyrate (see Alexiou & Franck 2008). Typical of prebiotics are inulin and oligofructose, both naturally present in a number of fruits and vegetables (e.g. bananas, chicory, Jerusalem artichokes, onions, garlic and leeks, and wheat), and other resistant oligosaccharides such as inulin-type fructans. Champ et al. (2003) also demonstrated a specific role for resistant starch in the stimulation of bacteria able to produce butyric acid. Human breastmilk contains complex oligosaccharides thought to be the principle growth factors for bifidobacteria (Cummings & Stephen 2007) and so is important in the development and maintenance of intestinal defences against pathogens. Regular consumption of prebiotics, such as oligosaccharides (FOS and inulin), is associated with reported improvements in the gut barrier function against infection. Most of the evidence is derived from animal studies; data from randomised clinical trials in humans exist but are generally less convincing (see Lunn & Buttriss 2007; Alexiou & Franck 2008). There may possibly be specific benefits for babies and elderly people in terms of acute infection and/or immune system effects but large trials are still required to determine whether these benefits exist (Elia & Cummings 2007). Recently, somewhat unexpectedly, both animal and human studies have reported increased rates of calcium absorption and associated improvements in bone mineral density with ingestion of prebiotics (see Alexiou & Franck 2008). Animal studies have revealed enhanced absorption in the colon of calcium, magnesium and iron with GOS, FOS and inulin, and five out of eight studies in humans also show a benefit, most importantly in adolescents (see Elia & Cummings 2007). It has been proposed that a change in ph associated with the production of SCFA is instrumental in promoting passive diffusion (see Alexiou & Franck 2008). This increase in absorption in the colon may counterbalance the potential for reduced absorption in the small intestine associated with consumption of high phytate cereals. Evidence is growing for a role for prebiotics in cancer protection (Elia & Cummings 2007), but again, most of the evidence is based on animal studies and further clinical trials in humans are required. Prebiotics are now being added to follow-on formulas for babies, based on evidence of amelioration of acute infectious diarrhoea and benefits in atopic disease (Fanaro et al. 2005), but large clinical trials are awaited. Furthermore, animal models have shown a reduction in blood cholesterol and triglycerides, although these findings are less consistent in human trials (Elia & Cummings 2007). With the exception of some studies in constipated subjects, oligosaccharides generally do not increase stool weight. The evidence emerging in the carbohydrates field and, in particular, work on prebiotics emphasises the importance of considering carbohydrate digestion in the context of an integrated whole gut process (Elia & Cummings 2007). Cancer Fibre intake has been linked to reductions in colorectal cancer risk. For example, the European Prospective Investigation on Cancer study reported a 40% reduction in risk of colorectal cancer between the lowest (15 g/ day) and the highest (35 g/day) quintiles of intakes (Bingham et al. 2003). In this study, the protective effect was reported for fibre from all sources; therefore, the type and the source of fibre may be irrelevant in terms of the benefits observed. It is worth noting that the average intake for adults in the UK is currently only about 13 g/day, and so a substantial increase will be required to meet the levels associated with protection in relation to cancer. The association with colorectal cancer has been supported by the World Cancer Research Fund (WCRF) who recently published the most comprehensive report on the relationship between cancer and diet, physical activity and weight (WCRF/AICR 2007). The report suggests that there is probable evidence (meaning the strength of the evidence is strong enough to generate public health recommendations) that foods containing dietary fibre (both foods naturally containing fibre, for example plant foods, and foods which have the constituent added) decrease the risk of colorectal cancer. Fruits and vegetables, sources of dietary fibre, are reported to be protective against other cancers, too. For example, the WCRF reported probable evidence that non-starchy vegetables and fruits are protective against cancers of the mouth, pharynx, larynx, oesophagus and stomach. Fruits were additionally found to be protective against lung cancer. The extent to which the fibre content of these foods contributes to protection is still unclear. Substantial evidence has specifically linked increased meat, particularly processed meat, consumption to colorectal cancer (see WCRF/AICR 2007 for an overview) and possible mechanisms have been summarised recently (Santarelli et al. 2008). Norat et al. (2005) demonstrated that the impact of high meat intake on risk can be offset by high intakes of dietary fibre (over 27 g/day), as illustrated in Figure 2.

10 Dietary fibre and health 195 Hazard ratio Diabetes High Medium Low Fibre intake High Medium Low Red and processed meat Figure 2 Relationship between red or processed meat consumption and fibre consumption. Source: Data extracted from Norat et al. (2005). Relative risk High (>78.6) Medium ( ) Low (<75.0) Low Medium High (<4.4 g/day) ( g/day) (>6.2 g/day) Cereal fibre intake Glycaemic index Figure 3 Relative risk of type 2 diabetes by different levels of cereal fibre intake and glycaemic index. Source: Data taken from Schulze et al. (2004). Some, but not all, cohort studies show an inverse association between total dietary fibre intake and risk of developing type 2 diabetes, but evidence is stronger for cereal fibre specifically, and a number of studies have also indicated a protective effect of whole grain foods (which typically have a low GI) (Venn & Mann 2004; Mann 2007). As shown in Figure 3, individuals consuming a diet with a high GI coupled with a low intake of cereal fibre have a risk of developing type 2 diabetes that is 75% greater than those consuming a high-cereal fibre intake combined with a low GI (Schulze et al. 2004). Mann (2007) commented that although diets rich in whole grains and dietary fibre may protect against diabetes and pre-diabetic states via promotion of satiety and weight loss, epidemiological data also provide support for an effect independent of fat mass. Evidence from intervention studies in subjects with type 2 diabetes supports the beneficial role of dietary fibre in improving glycaemic control, and this has been confirmed by a meta-analysis (Anderson et al. 2004).Viscous fibre (the soluble type present in oats and legumes) significantly reduced glycaemic response. Heart health Dietary fibre intake is inversely associated with risk of CHD (see Flight & Clifton 2006 and van Horn et al. 2008). In a pooled analysis of ten prospective cohort studies from the USA and Europe, Pereira et al. (2004) reported a 14% decrease in CHD risk for each 10 g/day increase in dietary fibre and a 27% decreased risk of coronary death. In reviewing the evidence for cardiovascular disease, Mann commented that although some analyses have pointed to a specific effect of cereal fibre, the relationship is not as strong as the inverse relationships between whole grain consumption and CHD, green leafy vegetables and CHD, or cruciferous vegetables, citrus fruits/juices and vitamin C-rich vegetables and fruits in the case of stroke (Mann 2007), suggesting that fibre per se is not necessarily the key factor. Perhaps not surprisingly, therefore, the mechanisms for a link between fibre intake and cardiovascular disease are unclear, but it is suggested that some constituents of dietary fibre, particularly those from fruits and grains (e.g. guar gum, pectin, beta-glucan and psyllium; so-called soluble fibres) can reduce blood cholesterol levels by altering cholesterol and bile acid absorption and by effects on hepatic lipoprotein production and cholesterol synthesis (see Lunn & Buttriss 2007). For example, a Cochrane Review found that diets rich in oats (a source of betaglucan) reduced low density lipoprotein (LDL) cholesterol by an average of 0.2 mmol/l compared with the control diet (Brunner et al. 2005). Energy balance and food control Diets low in energy and fat, as typically recommended for weight reduction, are poorly satiating but there is evidence that the addition of dietary fibre can help improve compliance. A recent review reported that an extra 14 g fibre/day can result in a 10% reduction in energy intake and an average weight loss of 1.9 kg in interventions lasting an average of 3.8 months (Slavin & Green 2007). Foods rich in dietary fibre tend to have a high volume and a lower energy density and may help to promote satiety (i.e. the state in which further eating is inhibited), decrease hunger, promote a sense of fullness

11 196 J. L. Buttriss and C. S. Stokes and hence, play a role in the control of energy balance (see Lunn & Buttriss 2007; Slavin & Green 2007). For instance, owing to their bulky nature, the increased consumption of cereals, fruits and vegetables has the potential to bring about a reduction in intake of other high energy density food, such as those high in fat. Results seem to differ depending on the fibre type and whether the fibre is added as an isolated fibre supplement rather than occurring naturally in a foodstuff (Slavin & Green 2007). NSP fibre (plant cell wall material) contributes to this effect, but the most effective at reducing subsequent energy intake, even at low amounts, are the viscous fibres (e.g. legumes, alginate, pectin, guar gum, psyllium and beta-glucan from oats and barley). These fibres also seem able to promote satiety perhaps because they are able to form a viscous gel in contact with water. In contrast, some non-viscous fibres, such as inulin and resistant starch, seem to have minimal effects even when consumed in large amounts (Slavin & Green 2007). Moreover, the processing of food determines its impact on satiety. For example, whole apples are more satiating than apple puree, which, in turn, is more satiating than apple juice. There is also some evidence that several mechanisms may be involved (e.g. effects on gastric emptying, transit time through the small intestine, gut hormone production) and that these mechanisms may be influenced to different extents by different types of fibre. For example, inulin-type fructans have been suggested to modulate appetite-regulating hormones (Cani et al. 2005), and oligofructose has recently been reported to induce satiety (Delzenne et al. 2007; Alexiou & Franck 2008). In contrast, resistant starches added to foods even at high doses may not be satiating (Slavin & Green 2007). Epidemiological studies indicate that fibre intake is linked to lower body weight although the impact is modest (van Dam & Seidell 2007). For example, in the Nurses Health Study, women who consumed more whole grains consistently weighed less than women who consumed less whole grain foods. Women in the highest quintile of dietary fibre intake had a 49% lower risk of major weight gain. van Dam and Seidell (2007) concluded that although results have been mixed, there is also some evidence that fibre supplements contribute to adherence to a low energy diet and hence weight loss. Food allergy and intolerance Some fibre-containing foods may need to be avoided in certain allergies. For example, wheat, rye and barley have to be eliminated from the diet in coeliac disease, which is thought to affect 1 in 300 to 1 in 100 people (Ellis 2002). Cereal peptides found in wheat (gliadin), rye (secalin) and barley (hordein) can trigger damage to the gut mucosa and consequently cause malabsorption of nutrients in susceptible individuals. Oats, provided they are uncontaminated with other cereals, need not be excluded from the diet. Some people suffer from specific fruit or vegetable allergies (Lessof 2002) and, in such cases, the trigger foods need to be avoided. Allergic reactions to fruits and vegetables (e.g. citrus fruit, kiwi, peach, celery) are usually mild, often affecting the mouth (oral allergy syndrome). In contrast, allergic reactions to peanuts (a legume) and to tree nuts can be very severe. Cooking, pasteurisation and other heat treatments typically reduce allergic potential (although this may not apply to peanuts). Ripeness can also make a difference in terms of allergic reactions. For example, the allergenicity of tomatoes increases with ripeness. There is also some evidence of possible immune system effects after the regular consumption of prebiotics (see earlier section). Fibre recommendations, labelling and health claims In the UK, the eatwell plate model (Fig. 4) is the basis for the government s practical advice on healthy eating. The two largest segments of the eatwell plate model describe the foods that provide most of the dietary fibre present in the food supply. Fibre can be obtained from the starchy carbohydrates group, which according to the model includes bread, potatoes (including low-fat oven chips), pasta, rice, oats, noodles, yams, maize, millet and cornmeal. The whole grain or wholemeal varieties, such as wholemeal bread or whole grain cereals, contain more fibre than the white varieties. Fibre can also be obtained from the fruits and vegetables group of foods and from pulses and nuts (that are grouped with meat, fish and eggs). Consumers also acquire information on healthy eating from food labels, associated advice based on derivatives of dietary reference values known in the UK as guideline daily amounts (GDAs), and from front of pack signposting schemes such as the Food Standards Agency s traffic lights scheme and the scheme based on GDAs. Currently in the UK, the debate surrounding the definition and measurement of dietary fibre has potentially resulted in inconsistencies in labelling. Arguably, it has also served to limit the emphasis given to fibre in various labelling schemes.

12 Dietary fibre and health 197 Figure 4 Eatwell plate model. Source: Food Standards Agency Crown copyright material is reproduced with the permission of the Controller of HMSO and Queen s Printer for Scotland. Available at: Fibre recommendations vary considerably around the world and the recommendation in the UK is among the lowest in the world. As has been described earlier, in the UK, the dietary fibre recommendation of 18 g/day is currently based on the fraction of non-digestible material described as NSP that is measured by the Englyst method. Yet, in line with the rest of Europe, food labelling information is typically provided in terms of the fibre content as analysed by the AOAC method, making comparison of intake data with the current dietary recommendation problematic. Although, as a rough approximation, dividing daily or weekly dietary fibre intake (as measured by AOAC) by 1.33 gives an approximation of NSP fibre (Englyst et al. 1996), the actual difference in the values provided by the two methods varies considerably from one food to another. Furthermore, there is currently potential for confusion about fibre content claims. The Regulation on Nutrition and Health Claims (European Commission 2007) allows claims to be made with respect to the fibre content of food if fibre levels exceed 3 g per 100 g (source of fibre) or 6 g per 100 g (high in fibre) (see Table 6). Yet, there is as yet no specific guidance on the methodology that should be used in this context (this is to be determined in the forthcoming revision to the food labelling legislation, which is currently at the proposal stage). The Regulation also provides a framework for health claims, both generally accepted claims linked to nutrient function (Article 13 claims) and disease risk reduction claims (Article 14 claims) (see Aisbitt 2007). Decisions on fibre-related claims are awaited, but in the meantime, it is of interest that outside the new EU regulation, Table 6 Framework for fibre content claims in Europe Source Increased High Fibre Either >3 g/100 g or >1.5 g fibre/100 kcal >30% more than a similar food for which no claim is made Either >6 g/100 g or >3 g fibre/100 kcal Note: Values quoted are assumed to be AOAC.