progress report dietary fibre - gut · may be five or six times the crude fibre intake, for, as...

Gut, 1973,14, 69-81

Progress report

Dietary fibre

Although an important component of animal feeds, fibre has for long beenconsidered an insignificant part of our diet. Recent advances in our knowledgeof its chemistry', coupled with some epidemiological2'3'4 observations haveled to the suggestion that it plays an essential role in the gut and in maintain-ing man's health. A relative lack of fibre in the diet of the 'Western' worldhas been linked with the increasing incidence of several common diseases5'6'7,and has provoked comment on the radio, television, and in the press8. Fibreis not an inert, unpalatable substance which passes unchanged through thegut, but a group of compounds with physical, chemical, and metabolicproperties which significantly alter gastrointestinal function.

Definition

Dietary fibre has been defined as that part of plant material taken in our dietwhich is resistant to digestion by the secretions of the human gastrointestinaltract9. It comprises a heterogeneous group of carbohydrate compoundsincluding cellulose and the hemicelluloses and a non-carbohydrate substancelignin. These substances form the structure of plants being present in the cellwalls of all parts including the leaf, stem, root, and seed. McCance andLawrence'0 called this fraction of the diet the 'unavailable carbohydrate',although it is not strictly unavailable or entirely carbohydrate.The term 'dietary fibre' is not synonymous with 'crude fibre', which is a

time-honoured and internationally accepted term used to define the fibrecontent of foodstuffs". Crude fibre, as defined by the association of OfficialAnalytical Chemists12 is the residue of a feeding material after treatment withboiling sulphuric acid, sodium hydroxide, water, alcohol, and ether. It is ameasure of the cellulose and lignin content mainly. The basis of this methodwas originally described by Hennenburg and Stohmann13 in 1860 and cameto be known as the Weende method after the town in Germany where itoriginated. However, it has been recognized for many years that this methodunderestimates the total amount of fibre in food by up to 50% 4,15. Thedifference between crude fibre and dietary fibre arises because during theWeende procedure the water- and alkali-soluble hemicelluloses are extractedand therefore not measured. Trowell has suggested that it is dietary fibreand not crude fibre that protects against disease.

Intake

In Britain crude fibre intake is between 4 and 8 g/day7"16 whilst in the UnitedStates intakes of about 8 to 11 g/day are reported17. True dietary fibre intakesare, however, very difficult to estimate as current tables of the compositionof food either do not give fibre content'8 or use the crude fibre value'9'20which gives only a relative idea of the total fibre content. Dietary fibre intakes

69

on January 23, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.14.1.69 on 1 January 1973. Dow

nloaded from

http://gut.bmj.com/

may be five or six times the crude fibre intake, for, as Southgate1 noted,whole wheat flour contains 11.2% of unavailable carbohydrate comparedwith only 2-3% crude fibre20. From the work of Southgate and Durnin21one may deduce that an unavailable carbohydrate intake, ie, dietary fibreintake, of between 16 and 28 g/day would be average in Britain. No surveysof dietary fibre intakes have been reported and true dietary fibre intakes areunknown. Variation within the community probably occurs as high crudefibre intakes are found in vegetarians17 (12-24 g/day crude fibre) especiallythose living in religious communities22 (22-35 g/day crude fibre). Total fibreintake is small compared with other components of the diet despite the factthat plants contain appreciable quantities of these substances. However,most of the fibre is removed from our bread and cereals by milling whilstpeeling and boiling reduce the intake from vegetables and fruit. A normalhelping of one of the 'high-fibre' breakfast cereals such as Kelloggs All-Brancontains enough fibre to double our daily intake.Much of the present interest in the role of dietary fibre in disease stems

from the suggestion that both the source and amount of our fibre intake haschanged during the past 100 years. The source has certainly changed. InBritain in 1835 flour consumption was around 360 lb per person per year23,by 1880 it had fallen to 280 lb24, and in 1970 to 145 lb16. Not only has flourconsumption fallen but it has been suggested that its fibre content has alsogone down over a similar period as a result of a change in the milling pro-cesses. In 1880 flour refining changed from stone-grinding to roller-milling.During the milling process flour is 'extracted' from whole wheat leavingbehind mainly bran. Stone-ground flour is an 80% extract of wheat and socontains more fibre than the roller-milled flour which is a 70% extract25.The impact of this change in the milling process on the fibre content of flourhas probably been overemphasized for, as Jones25 points out, flour which isan 80% extract has already had most of the fibre removed from it, furtherextraction depending on the removal of material other than fibre from theflour.At the same time as flour consumption fell so did our intake of potatoes

from 296 lb in 1880 to 224 lb per person per year in 197016. Concurrently,however, fruit and vegetable intakes rose and in fact doubled from 146 lbto 273 lb between 1909 and 197016,26. Similar changes have been noted forall these items in the USA, Norway, and Holland.26 Overall then, as Robert-son16 has suggested, total fibre intake may not have changed. Today we relymainly on fruit, vegetables, and nuts for fibre whereas bread and cerealproducts contribute only about 13%16. Our Victorian forbears probablyate less vegetables and more bread. These changes in fibre intake cannot beconsidered in isolation from alterations in the rest of our diet. Over this sameperiod equally significant changes have occurred in other components,including a striking increase in the consumption of both sugar3'27 and fat26.

Important geographical differences in fibre intake have been recordedand these form a further basis for interest in the role of dietary fibre indisease. Cleave, Campbell, and Painter3 noted a high fibre intake in ruralAfricans, such as the Zulus, most of their diet consisting of maize and beans.This has also been observed by Trowell2 and Walker28 and recently confirmedby Lubbe29. Lubbe compared crude fibre intake in 266 rural and 241 urbanVenda males by taking dietary histories, weighing the food, and performingsome chemical analyses. The mean crude fibre intake was 24.8 g/day in the

70 J. H. Cummings


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

rural population as against 5.7 g/day in the town. Apart from the Africanno other races have been so closely observed and no real attempt to quantitatedifferences in fibre intake at the international level has been made.

Chemistry30-33 (Table)

Dietary fibre is a mixture of high molecular weight polymers. The mainstructural units of the carbohydrate component are a variety of mono-saccharides whereas lignin is made up of phenyl-propane units. The amountof lignin present increases with the age of the cell (lignification) and is foundin close association with the carbohydrate components. Cellulose is the bestknown constituent of fibre and is the most abundant organic compoundfound in nature. It is composed entirely of straight chains of 1-4 ,B linkedD-glucose molecules which form fibres and take up water and swell. Thehemicelluloses were originally thought to be cellulose precursors but theyare not and are inappropriately named as they bear little structural relationto cellulose. They are a mixture of linear and highly branched polysaccharidescomposed most often of xylose units frequently substituted with galactose,glucose, mannose, arabinose, and other sugars. The B hemicelluloses, oracid hemicelluloses, contain uronic acids as galacturonic or glucuronic acidin addition to the hexose and pentose sugars. Sometimes classed with thehemicelluloses are the pectic substances. These include the plant gums andmucilages which are amorphous compounds as opposed to the other morefibrous polysaccharides. Pectin is an important constituent of the cell walland characteristically forms a gel on heating with sugars, as in jam making.

Fractionation and analysis of even the broad groups of constituents offibre is difficult and time consuming. A practical scheme for this which can becompleted for 'several samples' in 'just over five working days' has been setout by Southgate1.

Main Group Class Primary Unit

Carbohydrate Cellulose GlucoseHemicellulose A Xylose

GalactoseGlucoseMannoseArabinose

B 'Acid hemicelluloses' contain uronic acidsin addition

Pectic substances Mainly galacturonic acidNon-carbohydrate Lignin Phenyl-propane

Table Components of dietary fibre

Physiology

Dietary fibre is an unimportant source of energy for even if completelymetabolized in the gut the potential energy yield from 25 g would be only100 calories. However, fibre may affect the digestibility of other componentsof the diet since fat and nitrogen excretion have been shown to increase withhigh fibre intakes21. Fibre is metabolized by bacteria as it passes through thegut and is thought to exert a significant effect on faecal volume, laxation,

71Dietary fibre


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

intestinal transit time, flatus production, bacterial metabolism, and theoutput of organic anions such as bile salts and short chain fatty acids.

DIGESTIBILITYMangold34 in 1936 reviewing the digestibility of crude fibre pointed out thatalmost every animal including man had the capacity to digest fibre by virtueof his intestinal bacteria. Most bacteria and protozoa contain 'cellulases'which metabolize fibre to form short chain (or volatile) fatty acids (VFA),water, carbon dioxide, and methane. This process provides a significantenergy source to both ruminants35 and animals with a large caecum36, asthey are able to absorb the short chain fatty acids. Man possesses cellulose-splitting bacteria in his gut and excretes high concentrations of short chainfatty acids in his faeces37, but it is not certain to what extent he is able toabsorb the short chain fatty acids. Wide variations in the amount of fibredigested have been noted. The first major investigation in man was byWilliams and Olmsted38 (1936) who had earlier developed new methods forthe chemical identification of cellulose, hemicellulose, and lignin 4. In threemedical students they showed that on average 38 % (range 0-72%) of cellulose,56% (0-90%) of hemicellulose, and 35% (0-72%) of lignin disappearedduring passage through the gut, and that in general more hemicellulosedisappeared than cellulose, with lignin disappearing least. Using similarmethods these results were confirmed both in children39, who seem tometabolize fibre somewhat less, and in adults40. More recently Milton-Thompson41 (1971) has shown with careful balance studies that 57% (13-85%)of an 8.5 g/day intake of cellulose is metabolized during passage through thegut, and Southgate and Durnin21 found similar amounts of cellulose andhemicellulose being digested, with elderly subjects digesting more than youngadults. The apparent digestibility of lignin in the work of Williams andOmsted38 has been criticized on methodological grounds42. Lignin is probablyalmost indigestible42 and may well exert an inhibitory effect on the digestionof other components of fibre. Agronomists are aware that lignin reduces thedigestibility of forage crops, probably by its capacity to encrust fibre andform lignin-cellulose complexes43.

In man it has been suggested38 that the amount of lignin in food regulatesthe digestibility of cellulose and hemicellulose. The wide variation in theproportion of these substances digested probably depends on the relativeamounts of each present, the physical structure of the fibre, and on thevariable bacterial flora and time spent in the gut.

FAECAL OUTPUT AND BOWEL HABITIt has long been known44'45 that adding fibre-containing foods to our dietwill increase faecal bulk and improve bowel habit although few have actuallymeasured this. Cowgill and Anderson (1932)46 fed 1-1 1 oz of bran, eitheracid washed or unwashed, to five adults after a period on a low-fibre diet.The mean daily faecal weight during the low-fibre period was 114 g/dayand with bran 193 g/day. Similarly Williams and Olmsted38 fed an assortmentof high-fibre preparations such as wheat bran, alfalfa leaf meal, and carrotsto three medical students, and noted increases in daily faecal output of upto 100 g. More recently Eastwood et a147 have recorded significant increasesin the wet weight of stool in eight subjects when bran or cellulose was addedto their normal diet. In keeping with these experimental findings are the

72 J. H. Cummings


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

observations of Burkitt4 who noted daily faecal outputs of 470 g in 15African villagers living on a high-fibre diet, 185 g/day in Africans on amixed diet, and 108 g/day in senior boys at an English public school. Nomeasurements of intake were reported.

Others48'49 have observed qualitative differences with high-fibre diets suchas softening of the stool, more frequent bowel habit, and increased flatusproduction. Increased flatus has been noted after a high bean50 or brusselsprout51 diet, and Davies52 has measured greater flatus production in normalsubjects on a high-residue diet than in those on a no-residue intake. Theeffect of dietary fibre on intestinal gas composition is unknown.

TRANSITTransit time, the time taken for food residues to pass through the gut, isshortened by dietary fibre. A classic early study of this problem by McCance,Prior, and Widdowson53 showed in six normal people that brown bread(1.82% crude fibre) mixed with barium left the stomach and passed throughthe small intestine more rapidly, and left the colon 24 hours sooner than asimilar amount of white bread (0% crude fibre). Rural and urban Africanshave been studied with barium54, carmine55, and 'shapes'56. Walker56 noted38-62% of Bantus passed all their 'shapes' in 48 hours whereas only 16% ofthe group studied by Hinton et al57 of normal British male subjects did so.Burkitt's6 rural Africans passed 80% of their 'shapes' in a mean of 35 hourswhilst 24 young Englishmen took a mean of 77 hours to do so. Controlledstudies in Africans are not available, and although there is general agreementthat transit is more rapid in these people this has been ascribed to differencesin their diet other than fibre58. A diet rich in maize not only contains a lotof fibre but is relatively poorly digested and may hold large amounts of waterin the gut, thereby promoting more rapid transit.More recently in Britain Harvey et al59 studied eight patients with colonic

'stasis' and found a decrease in transit time (80% 'shapes') from 102 to68 hours after one to two months on a 'free' diet with added bran. No changein transit time was recorded in four subjects without stasis. Eastwood et al47have also reported no change in transit in eight normal subjects on theirusual diet with added bran or cellulose, despite an increase in faecal weight.Full details of these studies are not yet available.

VOLATILE FATTY ACIDSA clue to the way dietary fibre regulates faecal bulk and transit time may liein its relationship to volatile fatty acid production in the gut. Volatile fattyacids (acetic, butyric, propionic) are the major anions of normal faeces37,and at colonic pH they are 99 % ionized. As their lipid solubility is low theyare probably not absorbed and hence may act as osmotic cathartics. Theyare thought to be responsible for the diarrhoea of sugar intolerance60'61and some infant diarrhoeas62 and they may control faecal bulk in normalpeople63. Feeding dietary fibre has been shown to increase stool volatilefatty acid output at the same time as increasing faecal weight38. The originof volatile fatty acid in the stool is probably bacterial metabolism of un-absorbed carbohydrates, hence dietary fibre is a ready source of volatilefatty acids since a significant proportion of ingested fibre is metabolized.However, if faecal weight is increased simply by giving a cathartic such asmagnesium sulphate volatile fatty acid output will rise64, and in diarrhoea,

Dietary fibre 73


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

where faecal output varies for a number of reasons, the output increaseswith increasing stool weight (unpublished observations).

MECHANISM OF INCREASED FAECAL WEIGHTClearly the presence of 10 to 15 g of dietary fibre in the faeces is not going tohave a significant effect on normal stool volume which may vary from day today between 50 and 200 g. Fibre may exert its effect in one or all of three ways.The water-absorbing properties of fibre, particularly cellulose, are welldocumented. Fibre may act as a simple bulking agent and by virtue of theincrease in bulk promote colonic peristalsis and faster transit. However, asWilliams and Olmsted38 point out, this is probably not the whole explanation,for in their metabolic balance experiments each gram of fibre fed increasedthe stool volume in some studies by 12 to 20 g which seems more than couldbe accounted for by simple absorption of water alone. They noted also thatthe increase in faecal weight correlated best with the amount of dietary fibrebeing metabolized during passage through the gut. The main products ofthe bacterial metabolism of fibre are the volatile fatty acids, and as thesemay have a cathartic effect this could be an important factor in the increaseof faecal weight. Thirdly, the role of fibre in the small gut is largely unknownbut it is probable that substances such as bile salts and fatty acids aresequestered by the fibre. Since the fibre is subsequently metabolized in thecolon these compounds will be released and may themselves exert a catharticeffect65"6.

BILE SALTSThere is no general agreement as to whether dietary fibre binds bile salts inman and causes an increase in their excretion67. Lignin, in vitro binds di- andtri-hydroxy bile acids at acid pH probably by hydrophobic bonding68, butin man 10 g daily did not alter the half-life of labelled taurocholate or affectthe composition of the bile69'710. Lignin has been used successfully to controlthe diarrhoea of a patient who had had a massive intestinal resection7'but in five patients who had postileal-resection diarrhoea its effect was notsignificant statistically69. Adding cellulose to the diet of rats increases faecalexcretion of bile salts and their degradation products whilst shortening thehalf-life of cholic acid72. Turnover times are not always altered by cellulose,however73, although bile acid concentrations in the intestinal lumen are

increased74. In man the addition of cellulose to the diet has been shownto increase faecal bile saltoutput75-77 whilst methyl cellulose is withouteffect75. Studies using bran or varying the level of vegetable fibre intake havegiven conflicting results. For example, Eastwood et a147 were unable to showthat bran enhanced the excretion of bile acids. In a long-term and com-

prehensive study of white and Bantu prisoners in South Africa in 196278a crude fibre intake of 15 g/day was associated with significantly greaterbile salt outputs(3-21 g/week ±0-8 SD white: 2-75 g/week ± 0.45 SD Bantu)than a 4 g/day intake (2.24 g/week ± 0.7 SD white:2-21 g/week ± 0.5 SDBantu) whilst on a low fat diet. There was a simultaneous increase in faecalfat output. In the same study varying fibre intakes did not affect bile saltexcretion whilst on a high fat diet. Bile salt studies are still very method-dependent which probably explains the discrepancies in some of this work.Not only is it difficult to determine exactly which components of fibre bindbile salts in man, but also to differentiate the effects of fibre from those of

74 J. H. Cummings


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

bacterial binding of bile salts and from any concurrent changes in transittime and fat output.

Clinical Aspects of Fibre

The role that dietary fibre plays in the gut by increasing faecal bulk, shorten-ing transit time, and altering the metabolism of bile salts has been linkedwith epidemiological data to form a hypothesis which attributes the highprevalence of a wide group of diseases to lack of fibre in the diet. These'fibre-deficient' disorders include varicose veins, deep vein thrombosis andhaemorrhoids, diverticular disease of the colon, colonic cancer and polyps,appendicitis, constipation and the irritable bowel syndrome, diabetes, dentalcaries, atherosclerosis, and coronary artery disease.

CONSTIPATIONAs a simple remedy for constipation the laxative value of high fibre dietsis undisputed and has been the subject of many communications79-85.Furthermore various preparations of fibre form the basis of a large numberof laxatives. The mechanism by which fibre may increase faecal bulk hasalready been outlined, and this increase is well documented38 46'47. Cerealfibre in the form of bran may be more effective than fruit and vegetable intreating constipation80, which suggests that not all the components of fibrecontribute equally to this improvement in bowel habit. Lignin has been usedas an antidiarrhoeal agent69-7' and may counter the cathartic effects ofcellulose and hemicellulose.

DIVERTICULAR DISEASEMuch of the work relating diverticular disease of the colon to dietary fibrehas been done by Painter5'86'87. He points out that colonic diverticula aremost common in the sigmoid and left colon and are associated with hyper-trophy of the muscle and an increase in the segmenting pressures producedby the contracting colon88'89, and this has been confirmed by other workers90.Painter suggests that the increased pressure results from a fibre-deficientdiet which renders the colonic contents more viscous and thus harder topropel. The colonic lumen is narrower in these people because of the lowerbulk of the diet and so segments more efficiently, building up higher pressureswhich eventually produce diverticula.

Supporting evidence for the association of fibre and diverticular diseasehas been derived from epidemiological studies5 which contrast a prevalencefor diverticular disease of up to 45% in barium enemas and necropsy seriesin the 'developed' countries with an exceptionally low prevalence in ruralAfricans2,5,91. Less than a dozen cases have been observed in rural Africain the past 20 years and similar findings are reported from Korea92, Fiji,and Singapore93. The major difference between these areas of high and lowprevalence of the disease is reputed to be that of fibre intake. The changingpattern of dietary fibre intake in the UK already referred to may be a factorin the great increase in diverticular disease during this century. Deaths (permillion) rose from 2 to 16 (men) and 2 to 36 (women) between 1933 and19663. These death rates are not corrected for age and this is very much anage-related disease.

Rats94 develop diverticula of the colon when fed low-fibre diets. Those

Dietary fibre 75


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

rats which lived longest developed the most diverticula and had narrowercolons at necropsy. Hemicellulose in the diet (from psyllium seed husks) wasthought to offer the greatest protection against developing diverticula. Someconfirmation of this hypothesis has been sought by Painter et a195 who asked70 patients with symptomatic diverticular disease to take two teaspoonsfulof unprocessed bran (6-8% crude fibre) three times a day. Sixty-two patientsnoted improvement in their symptoms, including relief from pain, and moreregular bowel habit. The patients also changed to high fibre breakfast cereals,increased their fruit and vegetable intake, and cut down on their consumptionof sugar. Further experimental work relating colonic pressures, muscularhypertrophy, diverticula, and dietary fibre is needed but is hampered by thetime scale for the development of this disease.

COLONIC CANCEREpidemiological data have also been used to link dietary fibre with appendi-citis96, the irritable bowel syndrome97, and colonic neoplasms4'28'98. Burkitt4in a detailed survey of the incidence of colonic cancer notes that it is low inrural Africans (Kampala-3.5 per 100 000 population per year), high inEurope and the USA (Connecticut 51 8/100 000 per year), and that it changesin migrant populations such as American Negroes and Japanese living inHawaii. The geographical distribution of many colonic diseases is similarand suggests a common aetiology99"100 which is thought to be a low fibreintake in the diet. Fibre may protect against colonic cancer by decreasingtransit time and so allowing less time for carcinogens to be produced bycolonic bacteria and less time for the carcinogens to remain in the bowel4.However, the evidence for this theory is tenuous, for at the present time theavailable information on world fibre intake is not as well documented assome other branches of epidemiology. Further, as Drasar and Irving haverecently pointed out, the incidence of colonic cancer throughout the worldcorrelates best with high intakes of dietary fat and protein together withother indices of affluence, and poorly if at all with fibre intake101. The originof the colonic carcinogens is probably the bacterial metabolism of bilesalts and dietary fat. Professor Williams and his colleagues102 at the Wright-Fleming Institute in London have compared faecal bacterial flora from ruralUgandans, Indians, and Japanese where the prevalence of colonic cancer islow with the flora of English, Scottish, and American college staff andstudents who represent high colonic cancer areas. Higher counts ofbacteroideswere found in the stools of those living in high cancer incidence areas and agreater proportion of these bacteria were able to dehydroxylate bile salts.In addition their stools contained higher concentrations of faecal steroids andtheir metabolites. The low cancer risk group, particularly the rural Ugandans,are thought to have high fibre intakes. The role of fibre in altering bacterialflora and metabolism remains to be clarified.

VENOUS DISORDERSThe association between venous disorders and dietary fibre has been outlinedby Cleave3'103. He suggests that varicose veins, haemorrhoids, and venousthrombosis are produced by colonic pressure on the venous return from thelower limbs. Both the sigmoid colon and caecum are so closely related to thefemoral veins that when they become overloaded with dry, stagnating, low-residue faeces venous return is obstructed. Burkitt6 has modified this concept

76 J. H. Cummings


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

and suggests that the geographical variations in the incidence of thesedisorders is similar to that of the other fibre-deficient diseases, suggesting acommon aetiology.

ATHEROSCLEROSIS AND CORONARY ARTERY DISEASEAtheroma and ischaemic heart disease are probably the most prevalent groupof diseases to have been linked with a lack of fibre in the diet. Trowell7"104"105has recently reviewed the evidence relating them and suggests that a highfibre intake increases faecal steroid output and as this is the main catabolicpathway for cholesterol in man the blood cholesterol is lowered and thedevelopment of atheroma retarded. Not all the evidence supports thishypothesis. Rural Africans, whose high fibre intake has been referred to, doindeed have a very low prevalence of ischaemic heart disease2"106, whilst itis rising in the urban Bantu107. However, serum cholesterol levels do not alwaysshow the expected differences between rural and urban Bantu108"109 evenwhen they are maintained for long periods on controlled high and low fibreintakes"10. Clear differences have been recorded"""12 but have been relatedto genetic factors"ll or the amount of exercise taken"'. In India low serumcholesterol levels have been related to low fat intakes"14. Furthermore, otherareas in the world where there is a low incidence of heart disease"-5"'16 donot have such high fibre intakes. Trappist monks, who are lacto-ovo-vegetarians, have lower serum cholesterol levels (180 mg%) than theirBenedictine counterparts (225 mg %) who eat a more varied diet"7. Althoughfibre intakes are higher in the Trappists a notable lack of any difference inthe prevalence of vascular disorders in the two communities has been pointedout1"8"'19. Vegetarians in general have low serum cholesterol levels120"12' butEastwood22 was unable to lower the serum cholesterol levels further invegetarians fed additional fibre.

Investigations in animals into the cause of atheroma have shown that anassortment of feeds including oats'22, chow'23, wheat straw'24, alfalfa meal'25,rice bran'26, various gums and mucilages'27, and Indian grown grain'28reduce atheroma formation even when a high fat diet is fed. Standard ratchow pellets (high crude fibre content) when fed to conventional and germ-free rats produced a three times increase in faecal steroid output over a basalno-residue diet, whereas adding pure cellulose was without effect'29. Notall these investigations have been designed primarily to elucidate the role offibre which makes it difficult to separate its effect on cholesterol metabolismfrom any associated change in appetite, food intake, or additional fat andprotein.The interpretation of experiments in man is further complicated by the

multiplicity of factors which control cholesterol levels in the blood. Instudies where fibre has been fed as part of cereal or vegetable foods theintake and type of fat and carbohydrate has been altered simultaneously.Short-term studies on the whole suggest a hypocholesterolaemic effect forfibre77,13134. Rolled oats fed to 26 male volunteers reduced the serumcholesterol from 251 mg% to 223 mg% in three weeks'30. In another studymen on a high fat diet with added bengal gram (Cicer arietinum) hadcholesterol levels of 160 mg% compared with 206 mg% in 10 controls nottaking the added gram'3' with faecal bile salt excretions of 305 mg/day and236 mg/day in the two groups respectively. Pure cellulose has been fed tovolunteers in three studies47'77"135. In two of these47'135 a daily intake of 15 to

Dietary fibre 77


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

78 J. H. Cummings

16 g of cellulose was without effect on serum cholesterol but at 100 g/day"in 10 girls (age 10-12 years) cholesterol fell from 226 mg% to 170 mg% in10 days whilst faecal bile salt excretion rose. Two long-term studies of serumcholesterol and fibre have been done and give conflicting results131"110 thushighlighting some of the iproblems in ascribing changes in lipid metabolismto fibre intake.

Conclusion

Dietary fibre is an important component of our food. Its role in the gut hasbeen underestimated because of an incomplete knowledge of its compositionand inadequate techniques for the measurement of each constituent. Theeffects of fibre on small intestinal digestive functions have not been investiga-ted nor is it known whether fibre is capable of altering bacterial metabolismin the colon. The epidemiological evidence relating fibre to disease isfascinating but comes almost entirely from one small area of the world-rural Africa-and should be confirmed in other racial groups. Fibre has anestablished place in the treatment of constipation and could be of importancein the prevention of a wide range of colonic and cardiovascular diseases.This is clearly a developing area both for the gastroenterologist, his patient,and those interested in the physiology of the gut in both health and disease.

J. H. CUMMINGSMRC Gastroenterology Unit, Central Middlesex Hospital, London, NWIO

References

Southgate, D. A. T. (1969). Determination of carbohydrates in foods. II. Unavailable carbohydrates. J. Sci.Food Agric., 20, 331-335.

2Trowell, H. C. (1960). Non-Infective Disease in Africa. Arnold, London.3Cleave, T. L., Campbell, G. D., and Painter, N. S. (1969). Diabetes, Coronary Thrombosis, and the Saccharine

Disease, 2nd ed. Wright, Bristol.4Burkitt, D. P. (1971). Epidemiology of cancer of the colon and rectum. Cancer (Philad.), 28, 3-13.'Painter, N. S., and Burkitt, D. P. (1971). Diverticular disease of the colon: a deficiency disease of western

civilization. Brit. med. J., 2, 450-454.6Burkitt, D. P. (1972). Varicose veins, deep vein thrombosis and haemorrhoids: epidemiology and suggested

aetiology. Brit. med. J., 2, 556-561.'Trowell, H. C. (1972). Ischemic beart disease and dietary fiber. Amer. J. clin. Nlutr., 25, 926-932.'Sunday Times (1972). 20 August, p. 34.'Trowell, H. C. (1972). Crude fibre, dietary fibre and atherosclerosis. Atherosclerosis, 16, 138-140."°McCance, R. A., and Lawrence, R. D. (1929). The carbohydrate content of food. Spec. Rep. Ser. med. Res.

Counc. (Lond.), 135."Ministry of Agriculture, Fisheries and Food (1960). The Fertilisers and Feedingstuffs Regulations, S.I. No.

1165.2Kent-Jones, D. W., and Amos, A. J. (1967). Modern Cereal Chemistry, 6th ed., p. 564. Food Trade Press.3Hennenberg, W., and Stohmann, F. (1860 and 1864). Beitrage zur Begrfindung einer rationellen Futterung den

Wiederkaiser, Pts. I and 11. Brunswick.4Williams, R. D., and Olmsted, W. H. (1935). A biochemical method for determining indigestible residue

(crude fibre) in feces. Lignin, cellulose and non-water-soluble hemicelluloses. J. biol. Chem., 108, 653-666."Remy, E. (1931). Experimentelle Studien zur Biochemie und Biologie der Rohfaser. Biochem. Z., 236, 1-18."Robertson, J. (1972). Change in the fibre content of the British diet. Nature (Lond.), 238, 290-292.7Hardinge, M. G., Chambers, A. C., Crooks, H., and Stare, F. J. (1958). Nutritional studies of vegetarians.

11 1. Dietary' levels of fiber. Amer. J. clin. Nutr., 6, 523-525."McCance, R. A., atd Widdowson, E. M. (1960). The composition of foods. Spec. Rep. Ser. med. Res. Counc.

(Lond.), 297."Platt, B. S. (1962). Tables of representative values of foods commonly used in tropical countries. Spec. Rep.

Ser. med. Res. Coun. (Lond.), 302.2"Watt, B. K., and Merrill, A. L. (1963). Composition of Food, Raw, Processed and Prepared. (Agriculture Hand-

book, No. 8.) U.S. Department of Agriculture, Washington, D.C.'21Southgate, D. A. T., and Durnin, J. V. G. A. (1970). Calorie conversion factors. An experimental reassessment

of the factors ustd in the calculation of the energy value of human diets. Brit. J. Nutr., 24, 517-535.22Eastwood, M. A. (1969). Dietary fibre and serum-lipids. Lancet, 2, 1222-1225."3Lloyd, E. M. W. (1936). Food supplies and consumption of different income levels. J. Proc. agric. econ. Soc.,

4,89-120."Levi, J., Bourne, S., Brittain, Hancock, Jevons, and Fellows, F. P. (1882). Tn Report of the 51st Meeting ofthe

British Association for the Advancement ot Science, 1881, pp. 272-289. London."Jones, C. R. (1958). The essentials of the flour-milling process. Proc. nutr. Soc., 17, 7-15.


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

Dietary fibre 79

1"Greaves, J. P., and Hollingsworth, D. F. (1966). Trends in food consumption in the United Kingdom. WklRev. Nutr. Diet., 6, 34-89.

'7Hollingsworth, D. F., and Greaves, J. P. (1967). Consumption of carbohydrates in the United Kingdom.Amer. J. clin. Nutr., 20, 65-72.

"Walker, A. R. P. (1971). Diet, bowel motility, faeces composition and colonic cancer. S. Afr. med. J., 45,377-379.

2"Lubbe, A. M. (1971). A comparative study of rural and urban Venda males: dietary evaluation. S. Afr. med. J.,45, 1289-1297.

3"Wood, T. M. (1970). Cellulose and cellulolysis. Wld Rev. Nutr. Diet., 12, 227-265.3'Thomber, J. P., and Northcote, D. H. (1961). Changes in the chemical composition of a cambial cell during

the differentiation into xylem and phloem tissue in trees. I. Main components. II. Carbohydrate con-stituents of each main component. Biochem. J., 81, 449-455, and 455-464.

3"Schubert, W. J. (1965). Lignin Biochemistry. Academic Press, New York and London.33Horton, D., and Wolfrom, M. L. (1963). Polysaccharides. In Comprehensive Biochemistry, edited by M.

Florkin and E. H. Stotz, Vol. 5. Carbohydrates, pp. 185-232. Elsevier, London.34Mangold, E. (1934). The digestion and utilisation of crude fibre. Nutr. Abstr. Rev., 3, 647-656.3"Bryant, M. P. (1970). Normal flora-rumen bacteria. Amer. J. clin. Nutr., 23, 1440-1450."McBee, R. H. (1970). Metabolic contributions of cecal flora. Amer. J. clin. Nutr., 23, 1514-1518.37Wrong, O., Metcalfe-Gibson, A., Morrison, R. B. I., Ng, S. T., and Howard, A. V. (1965). In vivo dialysis of

faeces as a method of stool analysis. I. Techniques and results in normal subjects. Clin. Sci., 28, 357-375.3"Williams, R. D., and Olmsted, W. H. (1936). The effect of cellulose, hemicellulose and lignin on the weight of

the stool. A contribution to the study of laxation in man. J. Nutr., 11, 433-449.3"Hummel, F. C., Shepherd, M. C., and Macy, I. G. (1943). Disappearance of cellulose and hemicellulose from

the digestive tracts of children. J. Nutr., 25, 59-70.4°Sealock, R. R., Basinski, D. H., and Murlin, J. R. (1941). Apparent digestibility of carbohydrates, fats, and

'indigestible residue' in whole wheat and white breads. J. Nutr., 22, 589-596.4'Milton-Thompson, G. J., and Lewis. B. (1971). The breakdown of dietary cellulose in man. Gut, 12, 853-854."Crampton, E. W., and Maynard, L. A. (1938). The relation of cellulose and lignin content to the nutritive

value of animal feeds. J. Nutr., 15, 383-395."9Raymond, W. F. (1969). The nutritive value of forage crops. In Advances in Agronomy, edited by N. C.

Bradly, Vol. 21, pp. 1-108. Academic Press, New York and London."'Cogan, T. (1588). The Haven ofHealth or the Art ofProlonging Life by the Force ofPhysicke. Thomas Creede,

London."'Elyot, T. (1541). The Castel of Health, Corrected and Some Places Augmented, The Seconde Boke, Cap. 11, Of

Breade, p. 28.Th. Berthelet, London."Cowgill, G. R., and Anderson, W. E. (1932). Laxative effects of wheat bran and washed bran in healthy men.

A comparative study. J. Amer. med. Ass., 98, 1866-1875."'Eastwood, M. A., Hamilton, T., Kirkpatrick, J. R., and Mitchell, W. D. (1973). The effects of dietary supple-

ments of wheat bran and cellulose on faeces. Proc. Nutr. Soc., 32, 22A."'Hoppert, C. A., and Clark, A. J. (1945). Digestibility and effect on laxation of crude fiber and cellulose in

certain common foods. J. Amer. dietet. Ass., 21, 157-160.4"Walker, A. R. P. (1947). The effect of recent changes of food habits on bowel motility. S. Afr. med. J., 21,

590-596."5Steggerda, F. R., and Dimmick, J. F. (1966). Effects of bean diets on concentration of carbon dioxide in

flatus. Amer. J. clin. Nutr., 19, 120-124."'Kirk, E. (1949). The quantity and composition of human colonic flatus. Gastroenterology, 12, 782-794."Davies, P. J. (1971). Influence of diet on flatus volume in human subjects. Gut, 12, 713-716."3McCance, R. A., Prior, K. M., and Widdowson, E. M. (1953). A radiological study of the rate of passage of

brown and white bread through the aigestive tract of man. Brit. J. Nutr., 7, 98-104.54Campbell, G. D., and Cleave, T. L. (1968). Diverticular disease of the colon. (Letter). Brit. med. J., 3, 741."'Walker, A. R. P., and Walker, B. F. (1969). Bowel motility and colonic cancer. (Letter). Brit. med. J., 3, 238."Walker, A. R. P., Walker, B. F., and Richardson, B. D. (1970). Bowel transit times in Bantu populations.

(Letter). Brit. med. J., 3, 48-49.57Hinton, J. M., Lennard-Jones, J. E., and Young, A. C. (1969). A new method for studying gut transit times

using radioopaque markers. Gut, 10, 842-847.5"Westhuizen, J. van der, Mbizvo, M., and Jones, J. I. (1972). Unrefined carbohydrate and glucose tolerance.

(Letter). Lancet, 2, 719."Harvey, R. F., Heaton, K. W., Pomare, E. W., and Burwood, R. (1972). Colonic transit studies: Effects of

dietary fibre on transit time, and distribution of colonic contents in patients with stasis. (Abstr.) InProceedings of the 33rd Meeting of the British Society of Gastroenterology, Aviemore. (Printed in theSociety's Handbook).

6"Launiala, K. (1968). The mechanism of diarrhoea in congenital disaccharide malabsorption. Acta paediat.Scand., 57, 425-432. C

"'Weijers, H. A., Van de Kamer, J. H., Dicke, W. K., and Ijsseling, J. (1961). Diarrhoea caused by deficiencyof sugar splitting enzymes. Acta paediat. (Uppsala), 50, 55-71.

"Torres-Pinedo, R., Conde, E., Robillard, G., and Maldonado, M. (1968). Studies on infant diarrhea. II1.Changes in composition of saline and glucose solutions instilled into the colon. Pediatrics, 42, 303-311.

"3Fordtran, J. S. (1971). Organic anions in fecal contents, New EngI. J. Med., 284, 329-330.6"Grove, E. W., Olmsted, W. H., and Koenig, K. (1929). The effect of diet and catharsis on the lower volatile

fatty acids in the stools of normal men. J. biol. Chem., 85, 127-136."5Mekhjian, H. S., Phillips, S. F., and Hofmann, A. F. (1971). Colonic secretion of water and electrolytes

induced by bile acids: perfusion studies in man. J. cdin. Invest., 50, 1569-1577.6"Ammon, H. V. and Phillips, S. F. (1972). Fatty acids inhibit intestinal water absorption in man: fatty acid

diarrhea? Gastroenterology, 62, 717."7Heaton, K. W. (1972). Bile Salts in Health and Disease. Churchill Livingstone, Edinburgh and London."Eastwood, M. A., and Hamilton, D. (1968). Studies on the adsorption of bile salts to non-absorbed com-

ponents of diet. Biochim. biophys. Acta (Amst.), 152, 165-173.""Heaton, K. W., Heaton, S. T., and Barry, R. E. (1970). Comparison of two bile acid binding agents, chole-

styramine and lignin. In Advance Abstracts of the 4th World Congress on Gastroenteroloky, p. 447.


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

80 J. H. Cummings

7"Heaton, K. W., Heaton, S. T., and Barry, R. E. (1971). An in vivo comparison of two bile salt binding agents,cholestyramine and lignin. Scand. J. Gastroent., 6, 281-286.

7"Eastwood, M. A., and Girdwood, R. H. (1968). Lignin: a bile salt sequestrating agent. Lancet, 2, 1170-1172."Portman, 0. W., and Murphy, D. (1958). Excretion of bile acids and P-hydroxy-sterol by rats. Arch.

Biochem., 76, 367-376.73Gustafsson, B. E., and Norman, A. (1969). Influence of the diet on the turnover of bile acids in germ free

and conventional rats. Brit. J. Nutr., 23, 429-442."Eastwood, M. A., and Boyd, G. S. (1967). The distribution of bile salts along the small intestine of rats.

Biochem. biophys. Acta (Amst.), 137, 393-396.7"Stanley, M., Paul, D., Gacke, D., and Murphy, J. (1972). Comparative effects of cholestyramine metamucil

and cellulose on bile salt excretion in man. Gastroenterology, 62, 816."Stanley, M. M. (1970). Quantification of intestinal functions during fasting: estimations of bile salt turnover,

fecal calcium and nitrogen excretions. Metabolism, 19, 865-875.7"Shurpalekar, K. S., Doraiswamy, T. R., Saundaravalli, 0. E., and Narayana, Rao M. (1971). Effect of in-

clusion of cellulose in an 'Atherogenic' diet on the blood lipids of children. Nature (Lond.), 232, 554-555."Antonis, A., and Bersohn, I. (1962). The influence of diet on fecal lipids in South African white and Bantu

prisoners. Amer. J. clin. Nutr., 11, 142-155."Cheadle, W. B. (1886). A clinical lecture on the pathology and treatment of chronic constipation in childhood,

and its sequel, atony and dilatation of the colon. Lancet, 2, 1063-1064 and 1116-1118.8"Dimock, E. M. (1936). The treatment of habitual constipation by the bran method. MD Thesis, University

of Cambridge.8"Olmsted, W. H., Williams, R. D., and Bauerlein, T. (1936). Constipation: the laxative value of bulky foods.

Med. Clin. N. Amer., 20,449-459.'"Cleave, T. L. (1941). Natural bran in the treatment of constipation. (Letter). Brit. med. J., 1, 461.B"Walker, A. R. P. (1961). Crude fibre, bowel motility and pattern of diet. S. Afr. med. J., 35, 114-115."Jones, F. A., and Godding, E. W. (1972). Management o.f Constipation. Blackwell, Oxford.""Kramer, P. (1964). The meaning of high and low residue diets. Gastroenterology, 47, 649-652.""Painter, N. S. (1968). Diverticular disease of the colon. Brit. med. J., 3, 475-479.8'Painter, N. S. (1969). Diverticular disease of the colon: a disease of this century. Lancet, 2, 586-588.""Painter, N. S. (1964). The aetiology of diverticulosis of the colon with special reference to the action of

certain drugs on the behaviour of the colon. Ann. roy. Coll. Surg. Engl., 34, 98-119."Painter, N. S. (1970). Pressures in the colon related to diverticular disease. Proc. roy. Soc. Med., 63, Suppl.,

144-145.'0Arfdwidsson, S. (1964). Pathogenesis of multiple diverticula of the sigmoid colon in diverticular disease.

Acta. chir. scand., Suppl. 342."Cleave, T. L., and Campbell, G. D. (1966). Diets and diverticulitis. (Letter). Brit. med. J., 2, 50-51."Kim, E. H. (1964). Hiatus hernia and diverticulum of the colon. Their low incidence in Korea. New Engl. J.

Med., 271, 764-768.""Kyle, J., Adesola, A. O., Tinckler, L. F., and de Beaux, J. (1967). Incidence ofdiverticulitis. Scand. J. Gastroent.,

2, 75-80."Carlson, A. J., and Hoelzel, F. (1949). Relation of diet to diverticulosis of the colon in rats. Gastroenterology,

12, 108-115."Painter, N. S., Almeida, A. Z., ana Colebourne, K. W. (1972). Unprocessed bran in treatment of diver-

ticular disease of the colon. Brit. med. J., 2, 137-140."Burkitt, D. P. (1971). The aetiology of appendicitis. Brit. J. Surg., 58, 695-699."7Painter, N. S. (1972). Irritable or irritated bowel. (Letter). Brit. med. J., 2, 46."'Bremner, C. G., and Ackerman, L. V. (1970). Polyps and carcinoma of the large bowel in the South African

Bantu. Cancer (Philad.), 26, 991-999."9Burkitt, D. P. (1970). Relationship as a clue to causation. Lancet, 2, 1237-1240.°"Burkitt, D. P. (1969). Related disease-related cause. Lancet, 2, 1229-1231.°5Drasar, B. S., and Irving, D. (1973). Environmental factors and cancer of the colon and breast. Brit. J.

Cancer, in press.°"Hill, M. J., Crowther, J. S., Drasar, B. S., Hawksworth, G., Aries, V., and Williams, R. E. 0. (1971). Bacteria

and aetiology of cancer of large bowel. Lancet, 1, 95-100.13Cleave, T. L. (1959). Varicose veins. Nature's error or man's? Lancet, 2. 172-175.104Trowell, H. C. (1972). Fiber: a natural hypocholesteremic agent. Amer. J. clin. Nutr., 25, 464-465.°0Trowell, H. C. (1972). Dietary fibre and coronary heart disease. Rev. europ. Etude clin. biol., 17, 345-348.°"Shaper, A. G. (1970). In Atherosclerosis: Proceedings of the Second International Symposium, edited by

R. J. Jones, p. 314. Springer, Berlin and New York.7"Seftel, H. C., Kew, M. C., and Bersohn, 1. (1970). Myocardial infarction in Johannesburg Bantu. S. Afr.

med. J., 44, 8-12.508Nel, A., du Plessis, J. P., and Fellingham, S. A. (1971). A comparative study of rural and urban Venda males:

biochemical evaluation. S. AJr. med. J., 45, 1315-1317.°"Hathorn, M., Gillman, T., and Campbell, G. D. (1961). Blood lipids, mucoproteins and fibrinolytic activity

in diabetic Indians and Africans in Natal. Lancet, 1, 1314-1318.°"Antonis, A., and Bersohn, 1. (1962). The influence of diet on serum lipids in South African white and Bantu

prisoners Amer. J. clin. Nutr., 10, 484-499.551Truswell, A. S., and Mann, J. J. (1972). Epidemiology of serum lipids in Southern Africa. Atherosclerosis,

16, 15.29.12Bronte-Stewart, B., Keys, A., and Brock, J. F. (1955). Serum cholesterol, diet and coronary heart-disease.

Lancet, 2, 1103-1108.'3Walker, A. R. P. (1969). Coronary heart disease: are there differences in racial susceptibility? Amer. J.

Epidem., 90, 359-364.114Mathur, K. S., Wahi, P. N., Malhotra, K. K., Sharma, R. D., and Srivastava, S. K. (1959). Dietary fat,

serum cholesterol and serum lipid phosphorous, in different socio-economic groups in Uttar Pradesh.J. Indian med. Ass., 33, 303-309.

"'1Stout, C., Morrow, J., Brandt, E. N, Jr., and Wolf, S. (1964). Unusually low incidence of death from myo-cardial infarction. J. Amer. med. Ass., 188, 845-849.


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

Dietary fibre 81

"'Medalie, J. H., Neufeld, H. N., Riss, E., Groen, J. J., Kahn, H. A., and Bachrach, C. A. (1968). Variationsin prevalence of ischemic heart disease in defined segments of the male population of Israel. Israel J.med. Sci., 4, 775-788.

"7Groen, J. J., Tijong, K. B., Koster, M., Willebrands, A. F., Verdonck, G., and Pierloot, M. (1962). Theinfluence of nutrition and ways of life on blood cholesterol and the prevalence of hypertension andcoronary heart disease among Trappist and Benedictine monks. Amer. J. clin. Nutr., 10, 456-470.

"'McCullagh, E. P., and Lewis, L. A. (1960). A study of aiet, blood lipids and vascular disease in Trappistmonks. New Engl. J. Med., 263, 569-574.

"'Barrow, J. G., Quinlan, C. B., Cooper, G. R., Whitner, V. S., and Goodloe, M. H. R. (1960). Studies inatherosclerosis. III. An epidemiological study of atherosclerosis in Trappist and Benedictine monks:a preliminary report. Ann. intern. Med., 52, 368-377.

"'Hardinge, M. G., and Stare, F. J. (1954). Nutritional studies of vegetarians. II. Dietary and serum levels ofcholesterol. Amer. J. clin. Nutr., 2, 83-88.

"2'Ellis, F. R., and Montegriffo, V. M. E. (1970). Veganism: clinical findings and investigations. Amer. J. clin.Nutr., 23, 249-255.

5'Fisher, H., and Griminger, P. (1967). Cholesterol-lowering effects of certain grains and of oat fractions inthe chick. Proc. Soc. exp. Biol. (N. Y.), 126, 108-11 1.

's3Kritchevsky, D., and Tepper, S. A. (1968). Experimental atherosclerosis in rabbits fed cholesterol free diets;influence of chow components. J. Atheroscler. Res., 8, 357-369.

'"Moore, J. H. (1967). The effect of the type of roughage in the diet on plasma cholesterol and aortic atherosisin rabbits. Brit. J. Nutr., 21, 207-215.

"'Cookson, F. B., Altschul, R., and Fedoroff, S. (1967). The effect of alfalfa feeding on serum cholesterol andin modifying or preventing cholesterol-induced atherosclerosis in rabbits. J. Atheroscier. Res., 7, 69-81.

"Vijayagopalan, P., and Kurup, P. A. (1970). Effect of dietary starches on the serum, aorta and hepatic lipidlevels in cholesterol-fed rats. Atherosclerosis, 11, 257-264.

"7Fahrenbach, M. J., Riccardi, B. A., and Grant, W. C. (1966). Hypocholestdrolemic activity of mucilaginouspolysaccharides in white leghorn cockerels. Proc. Soc. exp. Biol. (N. Y.), 123, 321-326.

"'Devi, K. S., and Kurup, P. A. (1970). Effects of certain Indian pulses on tie serum, liver and aortic lipidlevels in rats fed a hypercholesterolaemic diet. Atherosclerosis, 11, 479-484.

'"Gustafsson, B. E., and Norman, A. (1969). Influence of the diet on the turnover of bile acids in germ freeand conventional rats. Brit. J. Nutr., 23, 429-442.

30de Groot, A. P., Luyken, R., and Pikaar, N. A. (1963). Cholesterol-lowering effect of rolled oats. (Letter).Lancet, 2, 303-304.

"'Mathur, K. S., Khan, M. A., and Sharma, R. D. (1968). Hypocholesterolaemic effect of Bengal gram; along term study in man. Brit. med. J., 1, 30-31.

"'Edwards, C. H., Booker, L. K., Rumph, C. H., Wright, W. G., and Ganapathy, S. N. (1971). Utilization ofwheat by adult man: nitrogen metabolism, plasma amino acids and lipids. Amer. J. clin. Nutr., 24,181-193.

"'Luyken, R., Pikaar, N., Polman, H., and Schippers, F. A. (1962). The influence of legumes on the serumcholesterol level. Voeding, 23, 447-453.

""Keys, A., Anderson, J. T., and Grande, F. (1960). Diet-type (fats constant) and blood lipids in man. J. Nutr.,70, 257-266.

"'Keys, A., Grande, F., and Anderson, J. T. (1961). Fiber and pectin in the diet and serum cholesterol con-centration in man. Proc. Soc. exp. Biol. (N. Y.), 106, 555-558.


http://gut.bmj.com

/G


nloaded from

http://gut.bmj.com/

progress report dietary fibre - gut · may be five or six times the crude fibre intake, for, as...

Documents