1203

13. Woodruff MFA. The interaction of cancer and host-its therapeutic significance. NewYork: Grune and Stratton, 1980: 111.

14. Centers for Disease Control. Update on acquired immune deficiency syndrome (AIDS)among patients with haemophilia A. Morbid Mortal Wkly Rep 1982; 31: 644-46.

15 Menitove J, Aster RH, Casper JT, et al. T-lymphocyte subpopulations in patients withclassic haemophilia treated with cryoprecipitate and lyophilized concentrates NEngl J Med 1983; 308: 83-86.

16. Editorial. AIDS and preventive treatment in haemophilia. N Engl J Med 1983, 308:94-95.

17 Ragni MV, Lewis JH, Spero JA, Bontempo FA Acquired-immunodeficiency-likesyndrome in two haemophiliacs. Lancet 1983; i: 213-14

18. Editorial. Acquired immunodeficiency syndrome Lancet 1983; i: 162-63.19. White GC, Lesesne HR. Haemophilia, hepatitis, and the acquired immunodeficiency

syndrome. Ann Intern Med 1983; 98: 403-04.20 Centers for Disease Control. Possible transfusion-associated acquired immune

deficiency syndrome (AIDS)-California. Morbid Mortal Wkly Rep 1982; 31: 651.21. Moran CJ, Mowbray J, Richards JDM, Goldstone AH. Plasmapheresis in systemic

lupus erythematosus. Br Med J 1977; i: 1573-74.22 Lockwood CM. Experience with plasmapheresis in glomerulonephritis and other

allergic diseases. In: Dau PC, ed. Plasmapheresis and the immunobiology ofmyasthenia gravis. Boston: Houghton Mifflin Medical Division, 1979; 175-85.

23 Dau PC, ed. Plasmapheresis and the immunobiology of myasthenia gravis Boston:Houghton Mifflin Medical Division, 1979.

24 Jones JV, Clough JD, Klinenberg JR, Davis P. The role of therapeutic plasmapheresisin the rheumatic diseases J Lab Clin Med 1981; 97: 589-98.

25 Birdsall HH, Brewer EJ, Rossen RD, Moake JL. Clinical improvement associated withhypocomplementemia following plasmapheresis. In: Dau PC, ed. Plasmapheresisand the immunobiology of myasthenia gravis. Boston: Houghton Mifflin MedicalDivision, 1979: 191-98.

26. Imbach P, Barandun S, d’Appuzo V, et al High-dose intravenous gammaglobulin foridiopathic thrombocytopenic purpura in childhood. Lancet 1981; i: 1228-31

27. Schmidt RE, Budde V, Schafer G, Stroehmann I. High dose intravenous

gammaglobulin for idiopathic thrombocytopenic purpura. Lancet 1981; ii: 475-7628. Fehr J, Hofmann V, Kappeler U. Transient reversal of thrombocytopenia in idiopathic

thrombocytopenic purpura by high-dose intravenous gammaglobulin. N Engl J Med1982, 306: 1254-58.

29. Paul LC, van Es LA, van Rood JJ, van Leeuwen A, Brutel de la Riviere G, de GraeffJ.Antibodies directed against antigens on the endothelium of peritubular capillaries inpatients with rejecting renal allografts. Transplantation 1979; 27: 175-79.

30. Hendriks GFJ, Schreuder GM Th, Claas FHJ, D’Amaro J, Persijn GG, Cohen B, vanRood JJ HLA-DRw6 and renal allograft rejection Br Med J 1983; i 85-87.

31. Hendriks GFJ, D’Amaro J, Persijn GG, Schreuder GMT, Lansbergen Q, Cohen B,van Rood JJ. Excellent renal allograft prognosis with DRw6 positive donors in theface of HLA-DR mismatches. Lancet (in press).

32. Lehner T. The relationship between human helper and suppressor factors to astreptococcal protein. J Immunol 1982; 129: 1936-40.

33. Matas AJ, Simmons RL, Kjellstrand CM, Buselmeier TJ, Najarian JS. Increasedincidence of malignancy during chronic renal failure. Lancet 1975; i 883-86.

34. Herr HW, Engen DE, Hostetler J. Malignancy in uremia- dialysis versus

transplantation. J Urol 1979; 121: 584-85.35 Gantt CL. Red blood cells for cancer patients. Lancet 1981; ii: 363.36. Francis DMA, Shenton BK. Blood-transfusion and tumour growth. evidence from

laboratory animals. Lancet 1981; ii: 871.37. Burrows L, Tartter P. Effect of blood transfusions on colonic malignancy recurrence

rate. Lancet 1982; ii: 662.38. Hodgson WJB, Lowenfels AB. Blood transfusions and recurrence rates in colonic

malignancy. Lancet 1982; ii: 1047.39. Jones PM Artificial blood. Br Med J 1983, i. 246-47.

Public Health

HEPATITIS B VIRUS INFECTION IN TWOGAMBIAN VILLAGES

H. C. WHITTLEK. MCLAUCHLAN

A. K. BRADLEYA. B. AJDUKIEWICZ

Medical Research Council Laboratories, Fajara, The Gambia

C. R. HOWARD A. J. ZUCKERMAN

Department of Medical Microbiology, London School of Hygiene andTropical Medicine

IAN A. MCGREGOR

Liverpool School of Tropical Medicine, Liverpool L3 5QA

were still positive in late 1980. 4 cases of primaryhepatocellular carcinoma out of 672 adults have been

diagnosed in the past five years. All 4 were in HBsAg carriers.

INTRODUCTION

IN the Gambia 1 in 50 men are estimated to die of primaryhepatocellular carcinoma.’ Here, as in other parts of theworld, the tumour is strongly associated with chronic

hepatitis B virus (HBV) infection. 80% of Gambians withprimary hepatocellular carcinoma carry hepatitis B surfaceantigen (HBsAg) in their blood (Ajdukiewicz A., personalcommunication). Although little is known of the

epidemiology of HBV infection in the Gambia, detailed andexcellent data are available from studies in neighbouringSenegal.2 Here HBV infection occurs early in life; 80% of thepopulation either are infected or have been infected by 7 yearsof age. The virus is transmitted either during the period ofbreast-feeding or at primary school entry at 6-7 years of age.Attack rates are high in young children, averaging 20% peryear, and result in a high HBsAg carriage rate. Horizontaltransmission either from mother to child or from child tochild appears to be the main pathway of infection. Thissupposition is supported by data from Liberia where child-to-child transmission was thought to be a major factor in thespread of HB infection. 3Various markers of HBV infection are useful in

epidemiological studies.4 The presence of HBsAg in theblood denotes that an individual is both infected and may beinfectious, especially if the hepatitis B e antigen (HBeAg) isalso present. HBsAg-positive individuals invariably haveantibody to the core component of the virus. This marker inthe absence of HBsAg also denotes recent or past infection,since most immune persons possess antibody to both core andsurface antigens. However a proportion of the infectedpopulation, varying from 1507o in young children to 30% inadults, may possess core antibody as the sole marker ofinfection, either past or present and another 10% only makeantibody to HBsAg.For various reasons we decided to study the epidemi-

ology of HBV infection in the two Gambian villages ofManduar and Keneba. First, local data were needed before

Summary The prevalence of hepatitis B virusinfection was markedly different in two

neighbouring Gambian villages. 62% of children in Manduaraged 2-4 years were infected whereas in Keneba, the othervillage, only 27% of this age-group were infected. However,in both villages few infants were infected-none under 6months of age and only 2 of 58 between the ages of 6 and 12months. Carriage of hepatitis B surface antigen (HBsAg) washigh, reaching a peak of 36% in the 5-9 age-group inManduar and 17. 6% in the 2-4 age-group in Keneba. 86%of all the children under the age of five who were HBsAg-positive also carried hepatitis B e antigen (HBeAg). Thisproportion fell to 17-6% for children aged 10-14 years andto 12.9% for mothers. Infection clustered in families,transmission from sib to sib being of major importance. Thechances of a child being an HBsAg carrier were

approximately 42% if an elder sib carried the antigen, 27% ifeither mother or father was a carrier, and 15% if neithermother or father was a carrier. There were 4 HBeAg-positivemothers who were highly infectious, since 10 of 11 of theirchildren became HBsAg carriers. Carriage of surface antigenlasted many years; 63% of those carrying the antigen in 1972

1204

a trial of HBV vaccine could be started. Second, unique long-term demographic data are available from these two villagesbecause all births and deaths have been registered over thepast 30 years. Thirdly, a survey of HBV infection had beenundertaken in these villages in 1972 that providedbackground and longitudinal data for the present survey. 6

PATIENTS AND METHODS

The two villages, Manduar and Keneba, are situated 8 km apart inthe West Kiang District of the Gambia along the river Gambia in theSahel savannah of Africa. Their society is patrilineal and

polygamous. The people, mainly Mandinka and Jola, are largelyMuslim. Staple foods are millet, sorghum, and rice. Breast feedingis universal, continuing for 18-24 months. Malaria, worms, andgastroenteritis reach a peak towards the end of the short wet season. 6In the nuclear family (ie, same mother and same father) children, upto 3 or 4 in number, sleep in their mother’s room, 1 or 2 being in herbed. Members of the extended family usually live in separatedwellings in the same compound or enclosed area. Accurate

demographic data for these two villages have been kept since 1951.5Crude death rates are high (average 24 - 7 per thousand in Manduarand 36 - 7 per thousand in Keneba for 1951-1975). Ages are knownaccurately for individuals born later than 1951 and annual surveyshave revealed the seasonal pattern and prevalence of infectiousdisease. Between March, 1972, and March, 1973, sera from 1317people of these two villages were tested for HBsAg byradioimmunoassay (’Ausria I’, Abbott): 174 (13%) were positivewith a peak in the 10-14 year age-group. 6The present survey took place in November, 1980. About half of

those living in Manduar and Keneba (populations 408 and 936)were under the age of 15. All mothers and their children under the

age of 15 were asked to participate in the survey. Attendance rangedfrom 7307o for those aged 10-14 to 100% for those aged 5-9. 63 ofthe 174 HBsAg-positive people in 1972 were also tested. 58 weremothers or fathers of the children surveyed and 5 were children aged10-14 years.Serum was collected by fingerprick and tested for HBeAg and

hepatitis B core antibody (HBcAb) by radioimmunoassay (Abbott).All sera were tested for HBsAg by reverse passivehaemagglutination (’Wellcotest’, Wellcome). A random sample of25 HBsAg-positive sera retested after addition of HBsAg-specificantibody were negative. Insufficient volumes precluded the testingof all sera for all markers of infection. Those tested were chosen

solely because sufficient serum was available for the assays.

RESULTS

Infection in Children under Five Years of AgeHBcAg was only tested for in those over 6 months of age

who were not carrying HBsAg in their blood. HBsAgprevalence was higher for both age-groups (table I) inManduar than in Keneba but the differences are not

statistically significant. In both villages few children aged 6months to 2 years had HBcAb but the prevalence of thisantibody in the 2-4 year-old children was 4707o in Manduarand 11 6% in Keneba (p<*0 - 001, X test). Assuming that oursample of 34 tested for HBcAb was representative of the 45

TABLE I-PREVALENCE OF HBsAg AND HBcAb IN CHILDREN LESSTHAN 5 YEARS OF AGE IN MANDUAR AND KENEBA

* Tested in children over 6 months of age who were HBsAg-negative.

TABLE II-AGE-SPECIFIC PREVALENCE OF HBsAg AND PROPORTIONOF HBsAg CARRIERS WHO CARRY HBeAg IN KENEBA AND MANDUAR

HBsAg-negative children aged 2-4 years in Manduar then62% (16 HBsAg-positive and 21 HBcAb-positive) of the 60children aged 2-4 in Manduar were infected with the virus asopposed to 27% of the 102 children in Keneba.There were few infections in children less than 1 year of

age. 1 in 18 in Manduar were HBsAg-positive and 1 in 40 inKeneba. None of the 28 sera of children under 6 months of

age was HBsAg-positive.

HBsAg and HBeAg in Children and Their Mothers

Prevalence ofHBsAg was high, reaching a plateau around 4to 6 years (table II). At all ages HBsAg prevalence was greaterin Manduar than in Keneba; the differences for each age-group were statistically significant from 5 years onwards(p<0 . 05, XZ test). The overall sex ratio for the HBsAg-positive children of the two villages was 50 males to 53females.The proportion of HBsAg-positive children who also

carried HBeAg was high in the youngest children and this fellwith age in both villages (p<0 0 1 for Manduar, p<0 02 for

Keneba, X2 test) (table n). 86% of the children (25 of 29 tested)under 5 who were positive for HBsAg were also positive forHBeAg and this fell to 17 6% (3 of 17 tested) for childrenaged 10-14 years (p<0 00 1, X test with Yates’ correction). 4of the 58 mothers in Manduar carried HBeAg but no mothersin Keneba were carriers of this antigen.

Relation of HBsAg Titre to Age and to Carriage of HBeAgYoung children had higher titres of HBsAg than older

children; means and SD (log2 units) were 13-4&plusmn;3-9 9 and

9’8&plusmn;3’3, respectively (p<0’01, t test) (see accompanyingfigure). Children who carried HBeAg had higher titres ofHBsAg than those who did not. Means and SD in log2 unitswere 13 - 0:t3’ 3 and 7 - &deg; 9&plusmn;3 1, respectively (p<0 01, t test).

Chronic Long term Carriage of HBsAgOf the 63 people who were HBsAg-positive in 1972, 40

(63%) were still carrying the antigen in late 1980. 35 of thesecarriers were adults, the other 5 were older children aged10- 14 years. 4 mothers in Manduar and 1 child also carried

HBeAg in their blood. Mean titre of HBsAg and SD in logzunits of 24 chronic carriers (all adults) was 9 - 2&plusmn;2 - 9 (seefigure).

Familial Clustering of HBsAg CarriersChildren carrying HBsAg clustered both in nuclear

families (p<0 &deg; 001, Xz test, only families with more than 2children analysed) and in extended families (p<0’001, test). The prevalence of HBs antigenaemia in children wasanalysed according to whether their parents were HBsAg

1205

Titres ofHBsAg in children and persistent carriers.

+HBeAg-positive; HBeAg-negative; 0 HBeAg status unknown.

positive or negative in 1972 (table III). Prevalence was higherin children of positive mothers than in those whose parentswere negative (p<0’ 01, X2 test). This association also held forchildren of HBsAg-positive fathers (p<0 &deg; 05, XZ test). 10 ofthe 11 children of the 4 HBeAg-positive mothers in Manduarwere carriers of HBsAg. ’

Prevalence of antigenaemia was also analysed according tosibship. If an older sib was currently positive for HBsAg thechances of his younger sibalso being positive was 42 - 207o(35/83), which was much higher than the figures of 9 - 9%(24/241) found in siblings in other families’in which the oldersib was negative for this antigen (p<0 01, X2 test). There were73 families with only 1 child, 9 of these children were HBsAg-positive, giving a prevalence of 12-3%, which is similar tothat in families in which -the older sib was negative forHBsAg. There was no association between family size and theprevalence of HBsAg carriers. Mean age and its range wassimilar in the various groups of children that were analysed.

Cases of Primary Hepatocellular CarcinomaSince 1976 a close surveillance for cases of primary liver

cancer has been kept in the two villages. 4 cases, proven eitherby liver aspirate and/or by a positive test for serum

alphafetoprotein, have been detected. 3 were men aged

TABLE III-PREVALENCE OF HBs ANTIGENAEMIA IN CHILDREN OF

PARENTS TESTED IN 1972*

* Combined data from both villages.

35-57, and 1 was a woman of 50. All 4 were carriers of

HBsAg at diagnosis and 2 were known to be carriers in 1972.The other 2 were not tested at this time.

DISCUSSION

We found that the prevalence of infection, measured bycarriage of HBsAg and presence of HBcAb, differed

. markedly between two neighbouring villages. Perhaps theseeds of HBV infection in Manduar, a village in which rateswere very high, were planted by 2 or 3 immigrants who werehighly infectious carriers of the virus or perhaps an itinerantmedicine man brought and spread the infection with his dirtyneedles. Certainly life styles, food, and eating habits weremuch the same in both villages yet crude mortality rates andthe prevalence of such important infections as those due toHBV and Plasmodium falciparum malaria were markedlydifferent. Local differences in the prevalence of commoninfections in African villages are often neglected and deservefurther study.HBV infection rates were high in both villages, rising

rapidly after the age of 2 years to reach a peak around schoolage. However, very few infants were infected. This findingsuggests that maternal antibody protects against HBV andthat, although a large proportion of mothers carried HBsAg,few mothers were so infectious that they transmitted the virusto their child during the perinatal period. Perhaps it is onlythe HBeAg-positive mothers (2 - 37o of all mothers) whotransmit the infection perinatally. We found that thesemothers were highly infectious- 10 of 11 of their childrenbecame carriers of HBsAg-but we do not know at what agethey infected their children. Prince et al drew similarconclusions from a survey in Liberia when they stated that"maternal transmission appears to be a minor factor in the

spread of hepatitis B infection in Liberia."3 These findingscontrast with those of Stevens et aF and Beasley et al8 inTaiwan where perinatal transmission by HBeAg-positivemothers is a major problem. These facts are of greatimportance in planning immunisation strategies, sincehorizontal infection in older children is much easier to

prevent by active immunisation than perinatal infection inyoung infants which requires both passive and activeimmunisation. Our data suggest that transmission of theinfection from sib to sib is of major importance in WestAfrica. If an elder sib was infected the chance of being apersistent carrier was 42 - 2%; this was higher than the ratesfound in families with an infected mother or father whichwere 28 . 201o and 26-6%, respectively. This high rate ofinfection correlates with the high prevalence of e

antigenaemia, a reliable indicator of infectivity,4 which wasfound in children in the two villages and which has also beenreported for nearby Senegal.’ Individuals with HBeAggenerally have high titres of HBsAg in their blood,8 whichexplains the high titres we found in the children, especiallythose under the age of 5 years. Contact between children in

bedroom, compound, or playground is close and although theprecise modes of transmission are not known many routes arepossible. Some have suggested mosquitoes;9 others

bedbugs, JO and recently we have found high titres of HBsAgin the exudates of tropical ulcers that are common in this area.Perhaps infectious sera from these sores is spread onto theskin of other children and inoculated by insects or bypenetrating thorns or sticks.Carriage of HBsAg persisted over long periods both in

Manduar and in Keneba; 63% of those positive in 1972 werestill carrying the antigen 8 years later. However, only 5 of

1206

these 40 chronic carriers, who were mainly adults, wereHBeAg-positive and their HBsAg titres were generally lowerthan those found in more recently infected children. On thisevidence many adult carriers may be less infective than e

antigen child carriers but they may experience the

consequences of chronic virus infection, such as cirrhosis andhepatoma, in later life. All 4 patients with primary livercancer carried HBsAg in their blood at diagnosis and 2 wereknown to have carried the virus for at least 7 years.This survey gives a guide as to the best age at which to use

hepatitis B vaccine in the Gambia. Infants should receive thevaccine around the age of 2 months when chances of infectionare low perhaps because maternal antibody is present. Suchyoung children are known to have a normal response to thevaccine" so this new and highly effective vaccine could begiven along with the diphtheria/pertussis/tetanus and oralpolio vaccines which are now widely and successfully givenduring the first 6 months of life in The Gambia. However,there is no proof that hepatitis B vaccine will lower theincidence of cirrhosis and hepatoma. This effect must beestablished before the widespread use of an expensive vaccinein a poor country is considered. HBV causes only subclinicalinfection in most of those affected.We thank the Chief of Manduar and Alkali Minte in Keneba for help. Mr K.

William Mr K. Jamneh, and Mr S. Rahman gave expert technical assistanceWe also thank Dr R. Whitehead, Dr A. Prentice and Dr B. M. Greenwood fortheir enthusiastic help and advice.

Correspondence should be addressed to H. C. W.

REFERENCES

1. Ryder RW. A proposal for a long term controlled trial of hepatitis B virus vaccine inThe Gambia to prevent hepatitis and liver cancer. Paper submitted to the GambianMedical and Health Department 1982.

2. Barin F, Perrin J, Chotard J, Denis F, N’Doye R, Diopmar I, Chiron J-P, Coursaget P,Goudeau A, Maupas P. Cross-sectional and longitudinal epidemiology of hepatitisB in Senegal. Prog Med Virol 1981; 27: 148-62.

3. Prince AM, White T, Pollock N, Riddle J, Brotman B, Richardson L. Epidemiology ofhepatitis B infection in Liberian infants. Infect Immun 1981; 32: 675-80.

4. Deinhardt F. Predictive value of markers of hepatitis virus infection. J Infect Dis 1980;141: 299-305.

5. Billewicz WZ, McGregor IA. The demography of two West African (Gambian)villages. 1951-75. J Biosoc Sci 1981; 13: 219-40

6. McGregor IA. Health and communicable disease in a rural African environment. Oikos1976, 27: 180-92.

7. Stevens CE, Beasley RP, Tsui J, Lee W-C. Vertical transmission of hepatitis B antigenin Taiwan. N Engl J Med 1975; 292: 771-74

8. Beasley RP, Trepo C, Stevens CE, Szmuness W. The ’e’ antigen and verticaltransmission of hepatitis B surface antigen. Am J Epidemiol 1977; 105: 94-98.

9. Prince AM, Metselaar D, Kafuko GW, Mukwaya LG, Ling CM, Overby LRHepatitis B antigen in wild-caught mosquitoes in Africa. Lancet 1972; ii: 247-53

10. Wills W, Larouze B, London WT, Millman I, Werner BG, Ogston W, Poutaghia M,Diallo S, Blumberg BS. Hepatitis B virus in bedbugs (Cimex hemipterus) fromSenegal. Lancet 1977; u: 217-19.

11. Barin F, Gourdeau A, Denis F, Yvonnet B, Chiron J-P, Coursaget P, Diomar IImmune response in neonates to hepatitis B vaccine. Lancet 1982; i: 251-53.

Nutrition: The Changing Scene

FERMENTATION IN THE HUMAN LARGEINTESTINE: EVIDENCE AND IMPLICATIONS

FOR HEALTH

JOHN H. CUMMINGS

Dunn Clinical Nutrition Centre, Addenbrooke’s Hospital,Trumpington Street, Cambridge CB2 1QE

THE principal functions of the large intestine are to

conserve water and electrolytes secreted into the gut duringdigestion, to provide a controllable route for the excretion ofwaste products of metabolism and of toxic substances, and tocontain safely the microorganisms that, by fermentation,conclude the digestive process. Fermentation, the anaerobicbreakdown of carbohydrate, is an important but hardlyrecognised part of human digestive physiology. It has a directbearing on salt and water absorption from the colon, on bowelhabit, on the excretion of toxic substances, and on nitrogenand sterol metabolism, and it may also influence intermediarymetabolism in the colonic epithelium, liver, and peripheraltissues.

EVIDENCE OF FERMENTATION

In animal physiology the. word fermentation is

conventionally used to describe the reaction in which

carbohydrate, usually polysaccharide, is broken down by theconcerted action of many species of anaerobic

microorganisms (see accompanying figure). The end-

products of this process are short-chain (or volatile) fattyacids, principally acetic, propionic, and butyric acids, and thegases carbon dioxide, methane, and hydrogen. Throughfermentation the microorganisms obtain energy for growth.What is the evidence that this process occurs in the humancolon?

Anaerobic MicrofloraThe large intestine of human beings contains a luxuriant

mixed culture of bacteria, 99% of which are anaerobes. Mostof these organisms are saccharolytic-ie, they derive theirenergy primarily from carbohydrate and its derivatives.2 Thenumber of organisms in colonic and faecal material has beenestimated at 10-10"/g, which means that we each havemore microbial than human cells. Difficulties in estimatingmicrobial size, however, and the extremely diverse microbialecology in the colon mean there is little information on theactual mass of microbial matter present. Early estimates putthe microbial mass at 25% of colonic solids, but more recentmethods indicate that it may be much higher, between 40%and 55% (ie, about 200 g microbial matter5). The

implications of the presence of such a large anaerobic organ inman have never been fully explored.

Available Substrates

The main carbohydrate passing from the ileum to thecaecum is the polysaccharide fraction of plant cell walls

(dietary fibre) together with some starch that may also escapedigestion. Only very small quantities of simple sugars are notabsorbed from the small intestine in healthy people, unlessthey are lactase deficient. The main source of endogenouscarbohydrate secreted into the colon is mucus. Intestinal

glycoproteins are 80% carbohydrate,6 and the human coloncontains microorganisms capable of degrading thesemolecules.’ 7

It is difficult to define what amounts of these substrates areavailable to the microflora. Average intake of the fermentablecomponent of dietary fibre is 15-20 g/day in the UK. Whenthis diet is eaten under controlled conditions 3-5 g of this

component appears undegraded in the faeces.8 The amountof starch that escapes breakdown in the small bowel is

currently being investigated. Using the evolution ofhydrogen in breath as a marker of colonic fermentation,Levitt and colleagues 9,10 estimate that 10- 15% of the starchin foods such as oats, white bread, and potatoes escapes