Caballero B. (ed.) Encyclopaedia of Food Science, Food Technology and Nutrition. Ten-Volume Set

Подождите немного. Документ загружается.

to their respective receptors along the intestinal wall

reducing absorption. The results of a recent meta-

analysis using several different clinical studies showed

a statistically significant lowering of plasma triacyl-

glycerides and cholesterol in subjects fed for at least

14 days with a diet supplemented with oats, psyllium,

pectin or guar gum, but the effect was modest: a

decrease of approximately 1 mg per deciliter per

gram of soluble fiber was observed. If we consider

that an individual with moderately higher blood

cholesterol has a total serum cholesterol value of

approximately 300 mg dl

1

, the change brought

about from consumption of viscous, soluble fibers is

on the order of 0.3%. Considering that oat bran

contains about 20% fiber by weight and only half of

this is soluble, one can calculate that 10 g of oat bran

per day is needed for each 0.3% change in total

cholesterol.

0014 Insoluble dietary fiber is, by definition, the fraction

of total dietary fiber (TDF) that is not soluble in hot

buffer solution. However, the measured proportion of

insoluble fiber can vary depending on the dietary fiber

methodology. Insoluble dietary fiber consists primar-

ily of hemicellulose, cellulose, and lignin (See Dietary

Fiber: Determination). It is often incorrectly believed

that insoluble fiber is not fermented to any great

extent by the bacteria present in the large bowel.

However, actual measurements of fiber fermentabil-

ity have shown that a significant proportion of insol-

uble dietary fiber is fermented. This is true even for

wheat bran, which has a relatively lower insoluble

fiber fermentability of between 30 and 40% (ap-

proximately 90% of the wheat bran TDF is insoluble

dietary fiber). The fermentability of oat bran is even

greater – up to 80% of the insoluble fiber (approxi-

mately half of the TDF) is fermented in the human

large intestine. The degree of fermentability is a prop-

erty unique to each dietary fiber and depends largely

on the nature and the structural arrangement of the

fiber components, and also on other physical charac-

teristics such as particle size. For example, the struc-

ture of coarse wheat bran favors a high water-holding

capacity (see below) and a slow rate of fermentation.

This permits the microtrapping of slowly released

fermentation gasses, increasing fecal bulk and stimu-

lating defecation.

Water-holding Capacity

0015 Dietary fiber holds water by adsorption and absorp-

tion. Some water is also retained outside the fiber

matrix (free water). The particle size, chemical com-

position, and structure of dietary fiber influence the

water-holding capacity. The in vitro water-holding

capacity cannot be used to predict the impact of

highly fermentable fiber on colonic function.

0016Although lettuce fiber isolates retain much water,

the water-holding capacity expressed in grams of

water held by fiber per gram of edible portion of

food has been found to be higher for wheat bran

fiber (4.5 g) and carrot fiber (2.1 g) than for that of

lettuce and various other foods (0.3–1.3 g). The

water-holding capacity of fiber from the brans of

ready-to-eat breakfast cereals is positively correlated

(r 0.85, P < 0.05) with its mean particle size (MPS):

the water-holding capacity of 160-mm MPS wheat

bran fiber is 59% of that of 800-mm MPS wheat bran

fiber (Table 2). A large portion of the water held by

wheat bran fiber appears to be free water.

0017The water-holding capacity has a significant effect

on fecal output and stool hardness, as discussed

above. These factors are important physiological

effects of wheat bran and contribute significantly to

its laxative effects.

Bile Salt Binding and Serum Cholesterol Reduction

0018There is considerable interest in the cholesterol-

reducing properties of dietary fiber. This is because

the influence of serum cholesterol (low-density lipo-

protein (LDL) and high-density lipoprotein (HDL)

cholesterol, in particular) on cardiovascular disease

is considerable. Three different mechanisms have

been proposed to account for the serum cholesterol-

reducing properties of dietary fiber. As discussed

above, a certain soluble viscous fiber may modestly

lower cholesterol by increasing luminal viscosity and

preventing cholesterol (re)absorption. Acetic and pro-

pionic acids, produced by the fermentation of dietary

fibers in the gut, are thought to inhibit liver choles-

terol synthesis at the metabolic level. Dietary fibers

also have the ability to bind cholesterol directly:

deconjugated bile salts are bound to pectic substances

by hydrogen bonding, and lignin appears to bind bile

salts through hydrophobic interactions. The actual

tbl0002Table 2 Effect of grinding on the properties of wheat bran

insoluble fiber

Mesh sieve Aperture

(mm) MPS (mm)

WHC

Glycocholate

binding

As is 800 9.5 + 0.1 26.5 + 1.5

20 840 420 8.1 + 0.1 24.9 + 1.2

40 420 280 6.6 + 0.2 23.8 + 0.8

60 250 180 5.8 + 0.1 22.1 + 1.0

80 175 160 5.6 + 0.1 21.5 + 0.3

Residue after neutral detergent and porcine pancreatic a-amylase

treatments.

Aperture size of sieve used with a Wiley mill, intermediate model.

MPS, geometric mean particle size.

WHC, water-holding capacity, grams of water per gram of insoluble fiber,

using the centrifugation method: mean + SEM of three measurements.

Grams of glycocholate bound per 0.2 g of insoluble fiber; mean + SEM of

four measurements.

1816 DIETARY FIBER/Properties and Sources

in vivo mechanism is unknown, and it is likely that

different fibers operate through different combin-

ations of all three mechanisms to influence blood

cholesterol levels.

0019 In vitro evidence suggests that the physical form of

the fiber may also play an important role in binding

bile salts. In commonly ingested ready-to-eat cereals,

glycocholate (r ¼0.90, P < 0.001) and taurocholate

(r ¼0.86, P < 0.05) binding have been positively cor-

related with the MPS of the neutral detergent fiber

(NDF; see Determination). Reducing the particle size

of wheat bran NDF from 800 to 160 mm reduces its

glycocholate binding by 19% (Table 2). The bile salt

binding capacity of fiber isolates, although method-

dependent, may reflect events in the terminal small

intestine. Purified fiber fractions may be poor

(cellulose) or strong (lignin) binders. Some rice

brans appear to have a high capacity to bind bile salts.

0020 Human cholesterol-lowering studies have linked

the consumption of dietary fiber with lower total

plasma cholesterol and increased LDL/HDL ratios

in addition to reductions in serum triglycerides. Stud-

ies such as these have suggested that cardiovascular

risk may be reduced by consuming particular types of

foods high in dietary fiber, most notably whole cereal

grains. However, it should be noted that coronary

heart disease (CHD) is a complex disease and is

affected by many factors. For example, the reduction

in CHD observed in a recent study was larger than

would be expected from the beneficial effects of sol-

uble dietary fiber on serum cholesterol levels only.

Thus, reductions in CHD are likely to be the result

of many factors, one of which is total dietary fiber

intake.

Cation-exchange Capacity

0021 Cation exchange is partly dependent on the presence

of uronic acid in the nonesterified form. The prepar-

ation of the fiber material may decrease the number

of nonesterified carboxyl groups and the apparent

cation exchange capacity. Wheat bran contains little

uronic acid, and its cation exchange capacity is

mainly due to diffusion within the dietary fiber net-

work. The cation exchange capacity of fiber from

different sources is difficult to compare, unless it is

expressed per edible portion. According to the affin-

ity of minerals for carboxylic acid groups, cabbage

and coarse wheat bran show a high cation-exchange

capacity compared with pectin.

0022 There has been some concern that the ability of

dietary fiber to bind minerals (as measured in vitro)

may lead to mineral deficiencies in individuals con-

suming high-fiber diets. In North American diets,

calculations show that mineral intakes far exceed

the potential binding capacity of dietary fiber so

that no need for concern arises. In addition, minerals

that are bound to fibers or that are trapped in the (as

yet, undigested) cell-wall matrix may not be absorbed

in the small intestine but could be partially released

and absorbed in the colon when the fiber is degraded

by bacteria. Absorption of minerals in the colon has

been suggested as a mechanism for accounting for

increased mineral absorption in rats fed fructooligo-

saccharides. This may explain why cation adsorption

has not been consistently related to mineral bioavail-

ability. However, wheat bran fibers can permanently

bind heavy metal ions to decrease their toxicity.

Viscosity and Gelling Properties

0023As mentioned above, certain soluble fibers, such as

oat b-d-glucans, are viscous when dissolved in water,

while others, such as pectins, show gelling properties.

These fibers influence gastric emptying and absorp-

tion rates in the small intestine. Direct measurement

of the viscosity produced by concentrations of fiber

likely to be used in diets has shown little effect of low

methoxy pectin, a slight increase with wheat bran,

and a significant increase with high methoxy pectin.

Some of the nonstructural dietary fiber components

can also increase viscosity.

Microbial Degradation (Fermentability)

0024Most dietary fiber remains undegraded until it reaches

the large intestine, where the extent of fermentation

depends on the source as well as several other factors,

including the physical structure of the fiber, the

presence of specific components in the fiber matrix,

the nitrogen source, bacterial adaptation, and transit

time. It is generally accepted that soluble fiber is

almost completely fermented in the large intestine,

but unfortunately, it is also generally believed that

insoluble fiber is not fermented. Actually, both

insoluble and soluble fiber are extensively fermented.

On average, 70–80% of the total dietary fiber from

mixed diets (e.g., fruits, vegetables, legumes, and

most cereals) is degraded by colonic bacteria. Since

insoluble dietary fiber is the major dietary fiber

fraction, representing approximately two-thirds of

total dietary fiber in Western diets, this means that

insoluble fiber is highly fermented. Indeed, as indi-

cated above, 70–80% of the insoluble fiber from oat

bran and up to 40% of the insoluble dietary fiber

from wheat bran are fermented.

0025There is some indication that fermentability

depends on particle size, with small particle sizes

being more readily fermented, but this has not been

consistently reported, and the effect appears modest.

In rats fed purified diets containing 15% hard red

wheat bran, NDF fermentability was largely un-

changed (34–36%) when the MPS varied from 1275

DIETARY FIBER/Properties and Sources 1817

to 394 mm. Similar trends were observed when Ameri-

can Association of Cereal Chemists soft and hard

wheat brans were fed at various MPS. In these experi-

ments, 37.3–37.6% and 30.0–32.8% of the NDF was

fermented, respectively.

Interaction between Structure and

Physiology

Food Processing

0026 The physicochemical properties of dietary fiber per se

are dependent on the chemical composition and the

structural characteristics of the fiber. However, as

noted above, knowing the former does not allow

one to predict the latter. When considering the poten-

tial in vivo effect of dietary fiber, one must also take

food-processing practices into consideration. Many

different types of food processing practices are

currently employed. Foods can be boiled, canned,

frozen, blanched, parboiled, extruded, adiabatically

extruded, and milled. All processes can have an effect

on dietary fiber content, as measured by current

methodologies. For example, heat treatment can

break glycosidic linkages in dietary fiber polysacchar-

ides. This can solubilize some of the insoluble fiber as

well as reduce the TDF content if the polymers are

broken down into small molecular mass fragments.

Heat treatment can also reduce the overall length of

the polysaccharide chains to lower viscosity and the

water-holding capacity of the fiber. The pentosans in

dietary fiber can react with amino acids such as lysine

(Maillard reaction) and form new polymers that do

not behave physiologically like dietary fiber. The

boiling of vegetables increases the apparent TDF con-

tent, but this seems to be due to the loss of nonfiber

components to the water.

0027 It is often difficult to predict before hand the effect

of a food-processing method on TDF and the distri-

bution between insoluble and soluble fractions. For

example, extrusion cooking of some cereal brans ap-

parently increases its water-holding capacity in vitro,

although this does not translate into a change in fecal

water holding capacity in vivo. Extrusion cooking of

wheat bran causes small decreases in TDF and insol-

uble dietary fiber but also gives rise to small increases

in soluble dietary fiber. Processing of rice (abrasion or

parboiling) removes the seed coat and the majority of

the dietary fiber. Some collapsing of the fiber struc-

ture also occurs during food preparation and/or mas-

tication due to the reduction of particle size, and the

fiber structure of sources such as spinach may be

modified by cooking.

0028 Milling is known to significantly affect TDF fer-

mentation in the human colon. Cereal grains have a

tough outer seed coat that would be completely resist-

ant to digestion if the cereals were not milled prior to

consumption. After milling, the thin-walled dietary

fiber structures from the endosperm are partly dis-

rupted and, thus, easily broken during fermentation.

Milling, therefore, makes the endosperm accessible to

the digestive enzymes in the upper digestive tract but

also makes the dietary fiber more accessible to co-

lonic bacteria. However, milling can also negatively

affect the physiological properties of the dietary fiber

in grains. While coarse grinding and adequate masti-

cation appear to be sufficient for digestion of cereals,

fine grinding disrupts much of the cell-wall architec-

ture, and further milling can even interfere with its

determination by gravimetric methods. Table 3 shows

the effect of milling on the dry bulk volume of wheat

bran, as measured in a volumetric cylinder. The

volume of the coarse bran was reduced from

5mlg

1

to less than 2 ml after grinding in a M5

Wiley mill using a 0.5-mm screen. Fiber from differ-

ent sources ground under the same conditions may

behave differently during milling, so that their

resulting particle sizes may be different.

0029As discussed above, the particle size of unfer-

mented wheat bran fiber may be related to colonic

function and fecal characteristics in humans and

laboratory animals. Finely ground cellulose or even

wheat bran, for example, may have a constipating

effect, whereas coarsely ground wheat bran prevents

constipation. The influence of particle size on fecal

bulking has been measured directly in rats. When rats

were fed diets containing 12% hard wheat bran, the

fecal wet density was 0.796 + 0.010 (mean + SEM,

n ¼12) in the coarse bran group (geometric MPS:

850 mm) but increased to 0.888 + 0.013 in the fine

bran group (MPS ¼308 mm). This increased density

was not due to a change in fecal weight but was due to

a decrease in fecal volume, paralleling the effect dem-

onstrated for bran in Table 3. In ready-to-eat break-

fast cereals, the fiber MPS has been found to vary

between 350 and 2000 mm. Increased particle size

normalizes gastrointestinal transit time, delays gastric

tbl0003Table 3 Effect of grinding on dry bulk of wheat brans (ml g

1

duplicate measurements

Sieve aperture

Bran A

Bran B Bran C Bran D

As is 3.1 3.4 5.1 3.5

2.0 mm (10 mesh) 2.7 2.4 2.8 2.1

1.0 mm (18 mesh) 2.2 2.1 2.4 2.0

0.5 mm (35 mesh) 1.8 1.8 1.9 1.7

Using a MS Wiley mill. Mesh sieve equivalence is indicated in

parentheses.

Bran A was American Association of Cereal Chemists-certified soft white

wheat bran; brans B–D were soft or hard wheat brans from other sources.

1818 DIETARY FIBER/Properties and Sources

emptying, and may also affect the extent of insoluble

fiber fermentation. Although dietary fiber has many

positive effects on the digestive tract, excessive

milling of dietary fiber to produce small particles in

the range of 7–70 mm may have negative effects.

Smaller particles do not exhibit the transit time nor-

malization associated with larger bran particles, and

the persorption of a small amount of very fine fiber

particles may occur in some regions of the intestinal

wall. Contrary to persorbed starch granules, per-

sorbed fiber particles are not degraded in the blood

and may impose an increased workload on the

kidneys. Data on persorption are scarce, and much

work is needed on this subject.

0030 Studies with apples, apple pure

e, and apple juice

have highlighted the link between the physical integ-

rity of the cell wall and the beneficial effects of un-

processed dietary fiber, in this case, a lowering of the

glycemic index of a food (the rise in blood glucose

that accompanies ingestion and digestion of a carbo-

hydrate-containing food). Only the unprocessed apple

had a reduced glycemic index; the apple pure

e, con-

taining the physically degraded dietary fiber was simi-

lar to apple juice in response. This demonstrates the

importance of the physical encapsulation of available

macronutrients within the fiber matrix. This effect

has also been observed with rolled oats where higher

glycemic indices were observed with more finely cut

oats. Studies such as these show that the integrity of

the cell wall is important in bringing about beneficial

physiological effects, even in the presence of viscous,

soluble fiber types (b-glucans in the cell wall; see

section Bile Salt Binding and Serum Cholesterol

Reduction).

In vitro

Measurements

0031 In vitro properties of fiber isolates have been used

to estimate their potential in vivo effect. However, in

vitro data should be used with caution. The measure-

ments are influenced by the methods used to follow

digestion in vitro as well as by the methods used to

prepare the fiber. Other confounding influences are

also present: the fiber residue obtained in vitro may

have been altered by heat or chemicals or may still

contain residual digestible materials. For these

reasons, in vivo conditions generally result in a more

extensive digestion than that measured in vitro.

Digestive Events

0032 An important consideration when examining the

physiological effect of dietary fiber is the fact that

dietary fiber is concentrated and denuded only in

the terminal small intestine. This means that the diet-

ary fiber that is partially associated with digestible

materials in the jejunum and duodenum may have

different physical properties from those of an entirely

denuded fiber matrix at the terminal ileum. In the

large intestine, most of the soluble fiber and a large

but variable portion of insoluble dietary fiber are

degraded by the colonic microflora. As already dis-

cussed, fermentation plays an important role in medi-

ating many of the physiological effects associated

with dietary fiber.

Definition (Based on Function and

Structure)

0033In the early 1970s, it became evident that cellulose

represents only a small part of the total dietary fiber

content of foods. Prior to this, there was a tendency to

equate dietary fiber with cellulose because that is

what was measured using the old crude fiber method.

While this helped reinforce the use of the term dietary

fiber (cellulose has a fibrillar structure), cellulose rep-

resents only a small part of the total dietary fiber

content of foods. Thus, at that time, new methodolo-

gies were emerging, and it was clear that there was a

need for a dietary fiber definition that was linked to

the actual physiochemical composition of dietary

fiber.

0034It is in this context that the dietary fiber hypothesis

appeared, a few years after Cleave’s hypothesis in

1956, which related refined foods and Western dis-

eases. The dietary fiber hypothesis thus appeared as a

refocusing of Cleave’s original ideas, reversing the

original emphasis to become: a diet high in unrefined

edible plant material is protective against Western

diseases of the bowel. The dietary fiber hypothesis is

important in understanding the ‘dietary fiber con-

cept,’ since it links botanical structure to digestive

processes and health outcomes – mainly noninfective

colonic diseases such as constipation, diverticular dis-

ease, irritable colon, ulcerative colitis, appendicitis,

hemorrhoids, polyps, and cancer of the large bowel. It

is this link to these and other health outcomes that

has characterized much of the work in the dietary

fiber field over the last 30 years. The first dietary

fiber definition in 1972 sought to describe the plant

structures and components that were responsible for

mediating these health benefits: ‘‘that portion of food

which is derived from the cellular walls of plants and

is digested very poorly by human beings . . . Fibre is

composed largely of cellulose, . . . hemicelluloses, pen-

tosans, pectin and lignin.’’

0035There are many different definitions of dietary fiber.

However, they all include certain specific and critical

aspects. First and foremost, a definition must include

the fact that dietary fiber corresponds mostly to the

structural material in the edible part of plant foods

in the usual human diet. This has been consistently

DIETARY FIBER/Properties and Sources 1819

reflected throughout decades by the inclusion of the

terms ‘plant cell wall’ when referring to dietary fiber

(Table 4). The term ‘dietary fiber’ itself is something of

a misnomer, since dietary fiber is not fibrillar (except

for cellulose). To reflect this fact, many other terms

have been proposed over the years, including unavail-

able carbohydrate and plantix. None of these terms has

succeeded in replacing ‘dietary fiber.’

0036 Another important aspect of any dietary fiber def-

inition is that it be resistant to human digestive secre-

tions (including enzymes) and not absorbed in the

small intestine. The observation that dietary fiber

increased fecal weight initially gave the impression

that dietary fiber was resistant to both human digest-

ive secretions and to bacterial digestion in the lower

gut (fermentation). While it is true that dietary fiber is

not digested and absorbed in the small intestine, a

large proportion of dietary fiber is fermented in the

colon. The extent of this fermentation is a function of

several different factors (see Microbial Degradation

(Fermentability) Section), meaning that the extent of

fermentation cannot be use as a criterion to define

dietary fiber.

0037Using ‘resistance to endogenous digestive secre-

tions’ alone as a criterion may be confusing. For

example, some dietary fiber may be partly degraded

by acids in the stomach, while some small intestine-

digestible material can reach the colon where bacter-

ial fermentation occurs. In addition, a definition

based on the resistance to digestibility by endogenous

digestive secretions could be problematic if the mater-

ial is not specified as coming from the edible part of

plant foods. Indeed, such a definition could be erro-

neously interpreted as meaning that dietary fiber is

any material resistant to in vivo digestion or, even

as any material measured as dietary fiber by a specific

dietary fiber method. Among the materials to be ex-

cluded from the dietary fiber definition are ‘ruminant

fiber’ and the products formed during processing and/

or cooking. There have been requests to include some

other plant components as well as manufactured

polymers, and this is a subject of much debate.

Among the criteria that could be used to accept or

reject a material as dietary fiber, only ‘edible plant

cell wall’ is specific enough to allow an absolute

definition. In addition, it links current dietary fiber

tbl0004 Table 4 Use of the term ‘plant cell wall’ in defining dietary fiber over time

Year Text Reference

1885 The ‘cell membranes’ of Rubner Trowell HC (1975) In: Burkitt DP and Trowell HC (eds) Refined

carbohydrate foods and disease, p. 43. Academic Press, London.

1929 The ‘skeletal framework of the plant’ McCance RA and Lawrence RD (1929) The carbohydrate content

of foods. Medical Research Council special report series

No. 135, pp. 1–173. HM Stationary Office, London.

1981 The ‘remnants of plant cells’ AOAC Fiber Consensus (1981) Association of Official Analytical

Chemists 95th Annual Meeting, October 1981, Washington, DC.

1985 ‘Derived from plant cell walls and not digested by human

alimentary enzymes’

Trowell H, Burkitt DP and Heaton K (eds) (1985) Dietary fibre,

fibre-depleted foods and disease, Academic Press, London.

1990 The ‘structural plant cell wall composed of

polysaccharides and lignin’

Eastwood M (1990) Function of dietary fibre in the large

intestine. In: Southgate DAT, Waldron K, Johnson IT and

Fenwick GR (eds) Dietary Fibre: Chemical and Biological

Aspects, pp. 211–219. Cambridge: Royal Society of

Chemistry.

1995 ‘The ‘‘cell wall material of plant foods’’ responsible for the

physical form and texture of unprocessed plant foods’

Heaton KW (1995) The dietary fibre concept – time for a

re-evaluation? In: Sørensen A, Bach Knudsen KE, Englyst HN,

Gudmand-Høyer E and Nyman M (eds) Metabolic and

Physiological Aspects of Dietary Fibre in Foods: Recent Progress in

the Analysis of Dietary Fibre, COST 92, p. 15. Luxembourg:

Commission of the European Commission.

1995 The ‘plant cell walls’ are the common characteristic of

plant foods that constitute a high-fiber diet associated

with beneficial effects.

Cummings JH, Hudson GJ, Quigley ME and Englyst HN (1995)

The classification and measurement of dietary carbohydrates.

In: Sørensen A, Bach Knudsen KE, Englyst HN, Gudmand-

Høyer E and Nyman M (eds) Metabolic and physiological aspects

of dietary fibre in food. Recent progress in the analysis of dietary

fibre in food, COST 92, pp. 17–36. Luxembourg: European

Commission.

1999 ‘the cell walls of edible plant tissues in the traditional

human diet’

Mongeau R, Scott FW and Brassard R (1999) Definition and

analysis of dietary fiber. In: Cho SS, Prosky L and

Dreher M (eds) Complex Carbohydrates in Foods,

pp. 305–316. New York & Basel: Marcel Dekker.

1820 DIETARY FIBER/Properties and Sources

definitions to the original ‘dietary fiber concept,’

which recognized that a diet high in minimally pro-

cessed edible plant foods provided many health bene-

fits. Confusion often arises because of the overlapping

characteristics and physiological effects of the differ-

ent ‘candidate’ materials with those of a dietary fiber

‘reference.’ Examples of ‘candidates’ are polydex-

trose (a randomly bonded synthetic glucose polymer),

inulin and fructooligosaccharides (naturally occur-

ring glucose-fructose oligomers and polymers), and

resistant starch (formed by food-processing methods

and largely undigested in the small intestine). This

overlapping means that it becomes virtually impos-

sible to distinguish between fiber and nonfiber mater-

ial based on resistance to digestion alone. Debates

such as this started shortly after dietary fiber became

popular in the 1970s. They are still ongoing, as dem-

onstrated by the publication of a recent international

survey of 147 professionals. While a complete discus-

sion of the merits of either side is beyond the scope of

this chapter, it is important to remember the origins of

the dietary fiber definition and the dietary fiber hy-

pothesis that firmly linked fiber to physiological out-

come.

0038 The definition of dietary fiber was first centered on

the ‘original dietary fiber concept’ that linked the

consumption of ‘plant cell walls’ in unrefined, trad-

itional foods to positive physiological outcomes.

Table 5 shows two definitions that were proposed

after the turn of the millennium. These two defin-

itions have taken the original ideas of the dietary

fiber hypothesis into account as well as included a

reference to some specific physiological effects in

humans.

0047 While the newer definitions of dietary fiber specif-

ically mention the fact that dietary fiber has physio-

logical benefits, the list is rather specific and deals

primarily with effects observed with viscous fibers

(lowering post-prandial serum glucose and lowering

serum cholesterol) or effects brought about from

partial fermentation and water holding capacity

(laxation). These effects are well supported by clinical

experiments with laboratory animals and with

humans. However, the original dietary fiber concept

related dietary fiber intake to the diseases of Western

man, including heart disease, cancer and obesity. As

mentioned above, the link between viscous dietary

fiber and heart disease is supported in the literature

through its effect on serum cholesterol and triglycer-

ides but the link with other Western diseases has been

much harder to demonstrate. It has been well estab-

lished that eating patterns that favor foods high in

dietary fiber and low in red meat consumption signifi-

cantly reduce the risk of cancer, heart disease and

obesity. However, it has been relatively difficult to

isolate this effect to dietary fiber consumption per

se. Part of the problem comes from the almost exclu-

sive reliance on epidemiological studies to demon-

strate any association. There are several reasons

why epidemiological studies are generally favored.

First, these studies provide data on free-living humans

rather than in-bred laboratory animals. Second, these

studies follow the long-term effect of dietary fiber

consumption on health outcomes. Third, the sample

size can vary from a few hundred (case control stud-

ies) to many thousands of individuals (prospective

cohort studies). This gives the studies good statistical

power and lends weight to the conclusions. However,

epidemiological studies also suffer from a number of

shortcomings including: problems with assessing

food intake for the past 2–3 years (case control stud-

ies), or with monitoring food intake 2–3 times over a

period of 5 years and using this to predict disease

outcome 15 years later (prospective cohort studies),

problems that arise from the use of incomplete dietary

tables to calculate dietary fiber intake, and problems

arising from the use of inappropriate methods for

correcting for energy intake.

0048The most striking conclusion from epidemiological

studies is a failure to provide support for an inverse

correlation between dietary fiber intake and colon

cancer. This concept has long been entrenched in the

literature largely because of studies with laboratory

animals that have provided much data in favor of

tbl0005 Table 5 Recent proposed definitions of dietary fiber

Year Text Reference

2001 ‘ . . . the edible parts of plant or analogous carbohydrates that

are resistant to digestion and absorption in the human small

intestine . . . promotes beneficial physiological effects including

laxation, and/or blood cholesterol attenuation, and/or glucose

attenuation.’

American Association of Cereal Chemists (2001) Report of

the Dietary Fiber Definition Committee to the Board of

Directors of the American Association of Cereal

Chemists. 2001. The definition of Dietary Fibre. Cereal

Foods World 46: 112–129.

2002 ‘Dietary Fiber nondigestible carbohydrates and lignin that are

intrinsic and intact in plants. Functional Fiber consists of

isolated, nondigestible carbohydrates that have beneficial

physiological effects in humans.’

National Academy of Sciences (2002) Food and Nutrition

Board. 2002. Dietary Reference Intakes. Proposed Definition

of Dietary Fiber. National Academy Press, Washington DC.

DIETARY FIBER/Properties and Sources 1821

a protective effect of dietary fiber. The animal studies

are supported by plausible physiological and bio-

chemical mechanisms that can account for the ob-

served results. However, the epidemiological data is

inconsistent in its support for an inverse relationship

between colon cancer and dietary fiber intake and

clinical trials that examined the effect of diet on the

re-appearance of colon polyps (colon cell growths

that have a 1/20 chance of forming a cancerous

tumor) found no protective effect of dietary fiber.

Even though the link between adenomatous polyps

and cancer is not straightforward, these studies show

that more research is needed to elucidate the relation-

ship between the known physiological benefits of

dietary fiber and cancer.

Sources of Dietary Fiber

0039 The total dietary fiber content of a food depends on

many factors, including (1) the plant variety, (2) the

stage of maturity when harvested, (3) the plant

growing conditions, and (4) the method of food

preparation. Although not all the methodological

issues have been completely resolved with respect

to measuring dietary fiber, there is now general con-

sensus on the total dietary fiber content for a majority

of foods.

0040 The names of 49 common dietary fiber sources

are provided in Table 6. The foods are classified

according to their TDF content expressed on an ‘as

is’ basis. Since some methodological variability is

inevitable, the letters ‘E,’‘M,’ and ‘P’ in the right-

hand column designate which methods would give

TDF values in the range shown. Other methods not

included in the comparison are likely to agree with

these values. The letter ‘E’ refers to NSP as measured

by the Englyst GLC procedure and includes a lignin

value measured separately with potassium perman-

ganate. The letter ‘M’ refers to the Mongeau rapid

method, and the letter ‘P’ refers to the Prosky method.

The last two methods are gravimetric and include

lignin in the final determination. Table 6 shows a

considerable agreement between methodologies: 33

of 49 foods are designated ‘EMP,’ meaning that

all three methods give similar values. In addition,

three foods are designated ‘MP’ and 13 foods as ‘M’

only because sample size limitations precluded

measurement with other methods. The TDF values

are based on winter and summer collections of up

to 20 individual foods over 30 months in different

locations.

0041 Most fruits and vegetables contain low amounts of

dietary fiber (1.0–2.2% TDF) on an ‘as is’ basis.

Legumes contain intermediate amounts of dietary

fiber (around 4% TDF), and cereals contain between

1–2.2% TDF (corn kernels) and 15.5–15.8% TDF

(oat bran). The dietary fiber content of foods is

greatly influenced by the moisture content. Thus,

cereal TDF values are high because these foods con-

tain little moisture (up to 10% at most), whereas

fruits and vegetables contain 80–90% moisture on

an ‘as is’ basis. The values of Table 6 are reported as

a percentage of weight, but when calculating dietary

fiber intake, it is important to factor in the serving

size (30 g for breakfast cereals, 50 g for lettuce and

raisins, and 125 g for vegetables and fruits). For

example, 1 serving of Bran flakes will provide ap-

proximately 4 g of dietary fiber, 1 serving of lettuce

will provide 0.35 g of TDF, 1 serving of raisins will

provide about 1.8 g TDF, and 1 serving of corn

will provide about 2 g of TDF.

tbl0006Table 6 Sources of dietary fiber

TDF

Foods Methods

0.1–0.5 White rice, cooked M

Orange juice from concentrate M

0.6–0.9 Lettuce M

Cucumbers EMP

Pineapple, melons (cantaloupe) EMP

1.0–2.2 White bead EMP

Corn kernel (canned), potatoes (boiled),

green beans (boiled), cabbage (raw),

celery (raw), onion (cooked),

green peppers (raw)

EMP

Cauliflower (cooked), tomatoes M

Apple, bananas, orange, pears,

strawberries, cherries, grapefruit,

plums

EMP

Apple pie, blueberry pie, pasta (cooked) M

2.3–3.0 Rye bread M

Mushroom (E ¼2.0) MP

Beets (canned), rutabaga (raw),

carrots (raw), broccoli (raw)

EMP

Blueberries (E ¼1.9) MP

3.1–4.3 Brown rice EMP

Raisins (seedless) M

4.4–6.4 Bran muffins, wheat cereal,

wholemeal bread

EMP

White beans (baked) EMP

Kidney beans (baked), peanut butter M

6.5–7.3 Peanuts EMP

Green peas (boiled) M

8.0–9.3 Oatmeal cereal EMP

9.4–10.3 Shredded wheat EMP

12.3–13.1 Bran flakes EMP

15.5–15.8 Oat bran EMP

40.6–40.9 AACC hard wheat bran (E ¼36.4) MP

TDF, total dietary fiber range in grams per 100 g on an ‘as is’ basis.

Methods: E, Englyst GLC method plus potassium permanganate lignin;

M, Mongeau rapid gravimetric method; P, Prosky gravimetric method.

From Mongeau R and Brassard R (1989) A comparison of three methods

for analyzing dietary fiber in 38 foods. Journal of Food Composition and

Analysis 2: 189–199; Mongeau R, Brassard R and Verdier P (1989)

Measurement of dietary fiber in a total diet study. Journal of Food

Composition and Analysis 2: 317–326.

1822 DIETARY FIBER/Properties and Sources

Glossary

.0042 Digestive secretions: enzymes and other secretions

produced by the digestive organs (e.g., salivary

glands and exocrine pancreas, stomach and small

intestinal mucosa, and liver); this excludes enzymes

produced by intestinal bacteria.

0043 Digestibility: degradation of ingested food com-

ponents by endogenous digestive secretions in the

upper gastrointestinal tract.

0044 Fermentability: degradation of undigested food

components (e.g., dietary fiber) by the intestinal

microflora, mostly in the large intestine.

0045 Pectic substances: mixtures of acidic and neutral

polysaccharides that can be extracted with water

from plant tissues and are rich in galacturonic acid;

together with hemicelluloses, they form the cell

wall matrix in which are enmeshed cellulose fibers.

0046 Persorption: the paracellular passage of small

amounts of solid food components from the lumen

of the intestine into the lymphatic and blood circula-

tion, this being without structural changes.

See also: Bile; Carbohydrates: Classification and

Properties; Cellulose; Dietary Fiber: Determination;

Physiological Effects; Effects of Fiber on Absorption;

Bran; Energy Value

Further Reading

Anderson JW, Allgood LD, Lawrence A et al. (2000) Chol-

esterol-lowering effects of psyllium intake adjunctive to

diet therapy in men and women with hypercholesterol-

emia: meta-analysis of 8 controlled trials. American

Journal of Clinical Nutrition 71: 472–479.

Baintner K (ed.) (1986) Intestinal Absorption of Macromol-

ecules and Immune Transmission from Mother to

Young. Boca Raton, FL: CRC Press.

Bolton RP, Heaton KW and Burroughs LF (1981) The role

of dietary fibre in satiety, glucose, and insulin: studies

with fruit and fruit juice. American Journal of Clinical

Nutrition 34: 211–217.

Brown L, Rosner B, Willett W and Sacks FM (1999) Chol-

esterol-lowering effects of dietary fiber: a meta-analysis.

American Journal of Clinical Nutrition 69: 30–42.

Carpita NC (1990) The chemical structure of the cell walls

of higher plants. In: Krichevsky D, Bonfield C and

Anderson JW (eds) Dietary Fiber: Chemistry, Physi-

ology, and Health Effects, pp. 15–30. New York:

Plenum Press.

Cummings JH (1986) The effect of dietary fiber on fecal

weight and composition. In: Spiller (ed.) Handbook of

Dietary Fiber in Human Nutrition, pp. 211–280. Boca

Raton, FL: CRC Press.

Dreher ML (ed.) (1987) Handbook of Dietary Fiber: An

Applied Approach. New York: Marcel Dekker.

Eastwood M (1990) Function of dietary fibre in the large

intestine. In: Southgate DAT, Waldron K, Johnson IT and

Fenwick GR (eds) Dietary Fibre: Chemical and Bio-

logical Aspects, pp. 211–219. Cambridge: Royal Society

of Chemistry.

Heaton KW (1995) The dietary fibre concept – time for a

re-evaluation? In: Recent Progress in the Analysis of

Dietary Fibre, Metabolic and Physiological Aspects of

Dietary Fibre in Foods, COST 92, p. 15. Luxembourg:

Commission of the European Communities.

Lee SC and Prosky L (1995) International survey on dietary

fiber: definition, analysis, and reference materials. Jour-

nal of AOAC International 78: 22–36.

Mongeau R and Brassard R (1982) Insoluble dietary fiber

from breakfast cereals and brans: bile salt binding and

water-holding capacity in relation to particle size. Cereal

Chemistry 59: 413–417.

Mongeau R, Scott FW and Brassard R (1999) Definition

and analysis of dietary fiber. In: Cho SS, Prosky L

and Dreher M (eds) Complex Carbohydrates in Foods,

pp. 305–316. New York & Basel: Marcel Dekker.

Selvendran RR and Robertson JA (1990) The chemistry of

dietary fibre: an holistic view of the cell wall matrix. In:

Southgate DAT, Waldron K, Johnson IT and Fenwick

GA (eds) Dietary Fibre: Chemical and Biological

Aspects, pp. 27–43. Cambridge: Royal Society of

Chemistry.

Spiller GA (1986) Definition of dietary fiber. In: Spiller GA

(ed.) Handbook of Dietary Fiber in Human Nutrition,

pp. 15–18. Boca Raton, FL: CRC Press.

Trowell H (1972) Ischemic heart disease and dietary fiber.

American Journal of Clinical Nutrition 25: 926–932.

Wolk A, Manson JE, Stampfer MJ et al. (1999) Long-term

intake of dietary fibre and decreased risk of coronary

heart disease among women. Journal of the American

Medical Association 281: 1988–2004.

Determination

R Mongeau, Ottawa, Aylmer (Que

bec), Canada

S P J Brooks, Ottawa, Canada

Introduction

0001This chapter presents a brief overview and compari-

son of commonly utilized methods for quantifying

dietary fiber in food. There are, fundamentally, only

two types of methods. Both rely on the removal of

nonfiber components by a combination of physical,

chemical, and enzymatic procedures prior to meas-

urement of dietary fiber. The gravimetric methods

measure dietary fiber by weighing the residue and

the gas-liquid chromatographic (GLC) methods

measure dietary fiber by hydrolyzing a treated sample

with acid to sugar monomers prior to quantification

DIETARY FIBER/Determination 1823

by GLC. Although, on the surface, both types of

methods appear similar, there are fundamental prac-

tical differences that have come about because of

variations in the individual method steps and differ-

ences in the choice of reagents and enzymes. These

practical differences have significant consequences

since they determine the final composition of dietary

fiber as measured by each method.

0002 Procedures have been described for isolating and

analyzing dietary fiber in specific foods as well as

analyzing carbohydrate polymers that are not

digested in the small intestine and are not identified

as dietary fiber by the current Association of Official

Analytical Chemists (AOAC) methods. The present

article, however, will focus on practical methods that

are applicable to the routine analysis of a wide range

of foods. Most total dietary fiber (TDF) methods also

allow for the quantification of different fiber frac-

tions such as soluble fiber, insoluble fiber, hemicellu-

loses, cellulose, lignin, etc. In this article, only the

methods that measure TDF are reviewed. This is

partly because of space limitations and partly because

the individual fiber components have not yet been

positively associated with human physiological

effects. For example, soluble and insoluble fiber can

be readily measured but these designations have little

physiological meaning since the digestive properties

and physiological benefits associated with some sol-

uble fiber can be found in some insoluble fibers and

the physiological benefits associated with insoluble

fibers are also found with some soluble fibers. In

addition, some processing techniques can also partly

depolymerize insoluble fiber that may then be found

in the soluble fiber fraction.

0003 The early work of Van Soest focused on measuring

ruminant fiber (mainly hemicellulose, cellulose, and

lignin) but these procedures could not be used to

measure TDF because they would have underesti-

mated the TDF content of human foods by missing a

significant proportion of the soluble dietary fiber. All

the modern methods incorporate steps to measure

soluble fiber. This is important since it can represent

a significant proportion of the TDF of a food. The

chemical complexity and physical variability of the

dietary fiber fraction in various foodstuffs as well as

the various food-processing techniques have con-

tributed to the difficulties encountered in devising

adequate dietary fiber methods. Many of the present-

day TDF procedures are based on the pioneering

work of Southgate.

0004 Ideal reference standards do not exist for dietary

fiber analysis. Purified fiber components and unpuri-

fied standardized ‘reference samples’ are both used as

standards of analysis, but the former group rarely

represents the complex dietary fiber matrix found in

foods and the latter group remains variable in com-

position, structure, and properties.

Historical Perspectives on Dietary Fiber

Values

0005Several different measures of dietary fiber and dietary

fiber fractions have been reported for foods over the

last 150 years. It is important to put these values into

some sort of context to understand the differences

between these measurements and the implications

for human dietary fiber research.

0006The crude fiber method was standardized in the

late 19th century in Germany. It was initially de-

veloped to analyze ruminant feeds and silage. It is a

highly empirical method that uses sequential extrac-

tions with hot dilute acid and alkali. The crude fiber

residue contains no soluble fiber and retains only

about 15% of the total hemicelluloses, 50–100%

cellulose, and 10–50% lignin. The variability in the

method and the variable losses of various dietary fiber

fractions mean that crude fiber values are of little

value in human nutrition. (See Cellulose; Hemicellu-

loses; Lignin.)

0007Acid detergent fiber (ADF) represented the first

method that reproducibly maintained specific dietary

fiber fractions. ADF is obtained by refluxing in

1 N mol l

1

sulfuric acid containing cetylmethylam-

monium bromide detergent followed by filtering and

washing. This fraction contains mostly cellulose and

lignin.

0008Neutral detergent fiber (NDF) represents the insol-

uble fraction of TDF. It is obtained by refluxing in a

solution of sodium lauryl sulfate followed by filtra-

tion and washing. NDF includes most of the cell wall

material but excludes the soluble b-glucans and sol-

uble hemicelluloses. It includes the ADF fraction as

well as the insoluble hemicelluloses. It thus under-

estimates TDF because it does not measure soluble

dietary fiber.

0009In some older tables, TDF values are reported as

the sum of NDF and pectin (estimated TDF) but, as

shown in Figure 1, this value is not always accurate.

Figure 1 also presents the relationship between insol-

uble dietary fiber (essentially NDF but using porcine

pancreatic a-amylase to remove starch; see below)

and TDF (as determined by the rapid HPB (Health

Protection Branch) method; see below) determined

for the same foods. The insoluble dietary fiber under-

estimates the TDF value by approximately 40% but

this is not constant. The differences between TDF and

insoluble dietary fiber represent actual differences

measured on the same food sample but the differences

between the estimated TDF and the actual TDF

values were performed at different times with

1824 DIETARY FIBER/Determination

different procedures and using different food samples

so that procedural variations may account for some

of the differences. For example, it is known that some

previously determined NDF values were measured

after incubation at lower (suboptimal) temperatures

that may have given higher and more variable values

due to incomplete starch digestion.

0010 In addition to the relationships described above,

Figure 1 shows the relationship between crude fiber

(CRF) and TDF (as determined by the rapid HPB

method) for foods from the Canadian Nutrient File.

Like the insoluble dietary fiber values, the CRF values

significantly underestimate the TDF values; however,

in this case, the variability is largely due to the CRF

method. Figure 1 shows that no consistent relation-

ship between TDF and CRF or insoluble dietary fiber

(and thus NDF). Thus, it is not possible to predict

TDF values from either CRF or insoluble dietary

fiber: a complete TDF measure is required.

Gravimetric Methods

0011 Gravimetric methods use enzyme and/or chemical

treatments to remove material that is digestible in

the small intestine. Ideally, these treatments should

mimic the digestive processes to arrive at a sample

that is representative of the material entering the large

intestine of humans. Because dietary fiber represents

the structural components of plant cell wall material

(See Dietary Fiber: Properties and Sources), the

weight of the residue is corrected for the presence of

noncombustible materials (such as minerals) and cor-

rections may also be applied for components that are

incompletely digested by the procedure (such as pro-

tein). The remaining weight is considered to be diet-

ary fiber. In order to measure the remaining material

gravimetrically, it must be filtered through glass cru-

cibles of defined porosities. It is, therefore, important

when preparing the sample to insure that it is not

too finely ground. The use of a Wiley mill avoids

these problems and minimizes particle loss during

filtration. Many different gravimetric methods have

been published with many variations in reagents,

temperatures and incubation times. However, rela-

tively few of these methods have been approved by

the Association of Official Analytical Chemists

(AOAC) after rigorous testing and interlaboratory

verification. Three AOAC approved methods will be

discussed in detail in this section: the rapid Health

Protection Branch method (rapid HPB method), the

Prosky method, and the Lee method. Before present-

ing these methods, we will first discuss the detergent

system.

The Detergent System

0012When introduced in 1963, the detergent system rep-

resented a great improvement over the CRF method

and it is still in use today for measuring insoluble

dietary fiber. It has been widely used to measure silage

and ruminant feed fiber. The system includes the NDF

and the ADF procedures, which use hot detergent to

remove digestible material and soluble fiber. The de-

tergent system discriminates among the insoluble

fibers – hemicelluloses, cellulose, and lignin. A che-

lating agent in the neutral detergent solution permits

the solubilization of insoluble pectin. When NDF and

ADF procedures are performed sequentially, the dif-

ference in weight between the NDF and ADF extracts

is an estimation of the insoluble hemicellulose frac-

tion. The ADF residue contains lignin, cellulose, and

(if present) cutin. Lignin can be measured by gently

oxidizing the ADF residue with potassium perman-

ganate. A 72% sulfuric acid treatment hydrolyzes

cellulose leaving cutin, phlobaphenes, and Maillard

products.

NDF and Porcine a-Amylase

0013Since water-insoluble dietary fibers represent a major

fraction of total dietary fiber, the NDF procedure

represents a simple method that can form part of an

024681012

TDF (rapid HPB method)

(g g

−1

'as is')

Estimated TDF, IDF, or CRF

(g g

−1

'as is')

fig0001 Figure 1 Relationship between crude fiber (CRF; white tri-

angles), insoluble fiber (IDF; gray circles), estimated total dietary

fiber (TDF: total of pectin þ neutral detergent fiber (NDF), black

circles), and TDF values for different types of foods. The TDF and

CRF values are from the Canadian Nutrient File (Health Canada

1997) and Mongeau et al. (1989). TDF and IDF values were deter-

mined by the rapid HPB method (AOAC 992.16, see text). The solid

line represents a 1:1 correspondence. The dashed line (IDF) and

the dotted line (CRF) are regression lines with the intercept fixed at

the origin. Insoluble dietary fiber is essentially NDF but includes a

treatment with porcine pancreatic a-amylase. The NDF þ pectin

values are ultimately from the USDA Nutrient Database (1999).

DIETARY FIBER/Determination 1825