Correspondence
Dietary Fiber and Colorectal Cancer
N Engl J Med 1999; 340:1924-1926June 17, 1999
- Article
To the Editor:
The fundamental flaw in the article by Fuchs et al. (Jan. 21 issue)1 concerns the definition of what constitutes a high-fiber diet. I was lucky enough to attend one of Dr. Denis Burkitt's lectures in London in 1972 and vividly remember how he defined a high-fiber diet. It had nothing at all to do with the calculations of Southgate et al.2 or any other calculation; whether a diet was considered to be high in fiber depended on the effect the diet had on the stool.
Burkitt began his lecture with a slide showing a stool of a typical Western, “civilized” person, a sausage-shaped thing, familiar to most of us, in a toilet bowl. His next slide was of a stool of a typical rural West African, which looked like a flat cow pie. Burkitt postulated that the West African's stool moved more quickly through the colon, giving carcinogens contained on its surface less time to be in contact with the mucosa — thus less time to induce carcinogenesis.
During the question-and-answer period, many questions from the audience concerned how one determined whether or not a diet was high in fiber in the sense Burkitt meant. Burkitt shook his head at all the salads, cereals, and breads offered as sources of fiber. He showed a slide of the staple cereal eaten by West Africans, which looked, in its wooden bowl, not too different from the stool that came out the other end. The only thing the study by Fuchs et al. proves is what anyone who heard Burkitt's lecture already knew: the American public has been sold a sugar-coated misconception.
2 ReferencesNeil D. Ravin, M.D.
5530 Wisconsin Ave., Chevy Chase, MD 208151
Fuchs CS ,Giovannucci EL ,Colditz GA , et al. Dietary fiber and the risk of colorectal cancer and adenoma in women. N Engl J Med 1999;340:169-176
Full Text | Web of Science | Medline2
Southgate DA ,Bailey B ,Collinson E ,Walker AF . A guide to calculating intakes of dietary fibre. J Hum Nutr 1976;30:303-313
Medline
To the Editor:
Fuchs and colleagues suggest little role for fiber in the prevention of colorectal cancer. A similar finding was reported in a British study of vegetarian diets.1 One limiting factor in both studies was that the subjects were adults. Colorectal carcinogenesis is a prolonged process and occurs over a period of several decades, even in patients with familial adenomatous polyposis or hereditary nonpolyposis colorectal cancer.
The incidence of colorectal cancer in India is one of the lowest in the world.2 The prevalence of adenomas is also very low. Our preliminary studies, however, revealed that the onset of adenoma or cancer was not delayed in Indian patients with inherited colorectal cancer syndromes, who were eating a vegetarian diet from early childhood.3,4 This finding suggests that a genetic predisposition cannot overcome the protective effect of fiber and a vegetarian diet.
4 ReferencesK.M. Mohandas, M.D.
Tata Memorial Hospital, Parel, Mumbai 400012, India1
Key TJ ,Thorogood M ,Appleby PN ,Burr ML . Dietary habits and mortality in 11,000 vegetarians and health conscious people: results of a 17 year follow up. BMJ 1996;313:775-779
CrossRef | Web of Science | Medline2
Mohandas KM, Desai DC. Epidemiology of digestive cancers in India: V. Large and small bowel. Indian J Gastroenterol (in press).
3
Mohandas KM ,Dhir V ,Jagannath P ,DeSouza LJ . The effect of life long vegetarian diet on familial adenomatous polyposis (FAP). Am J Gastroenterol 1998;93:1693-1693 abstract.4
Mohandas KM ,Swaroop VS ,DeSouza LJ ,Jagannath P . Hereditary nonpolyposis colorectal cancer in Indian patients. Indian J Gastroenterol 1996;15:Suppl 1:28-28 abstract.
Medline
To the Editor:
We would like to ask Fuchs et al. about the quality of their data on fiber, the potential mechanisms involved, and their view of carbohydrates in the diet.
The statistical analysis of the data is comprehensive, but the basic values for fiber may not merit such analysis. The data on fiber are drawn from a 1976 article by Southgate et al.,1 which gives fiber values for a limited number of foods in the United Kingdom. These foods are now known to contain some starch, and thus the method of analysis substantially overestimates the true intake of polysaccharides from the cell walls of plants. It would be valuable to see the data reanalyzed with nonstarch polysaccharides as the basis for fiber intake.2
Fuchs et al. adjusted all their data for energy intake. This adjustment may not be legitimate for fiber, and the authors should present absolute values for fiber intake.
In relation to colorectal cancer and adenomas, the evidence indicates that fiber works only if it is undigested in the upper gut and then fermented by anaerobic bacteria in the colon. This process produces short-chain fatty acids and alterations in the metabolism of nitrogen, bile acids, and bacteria. Other carbohydrates, such as resistant starch, the nondigestible oligosaccharides, some sugar alcohols, and in some populations, lactose, also have these properties.3 To understand the role of fiber in colorectal cancer we must account for these other dietary carbohydrates.
3 ReferencesJohn H. Cummings, F.R.C.P.
Addenbrooke's Hospital, Cambridge CB2 2QQ, United KingdomDavid A.T. Southgate, Ph.D.
8 Penryn Close, Norwich NR4 7LY, United Kingdom1
Southgate DA ,Bailey B ,Collinson E ,Walker AF . A guide to calculating intakes of dietary fibre. J Hum Nutr 1976;30:303-313
Medline2
Englyst HN ,Quigley ME ,Hudson GJ ,Cummings JH . Determination of dietary fibre as non-starch polysaccharides by gas-liquid chromatography. Analyst 1992;117:1707-1714[Erratum, Analyst 1993;118:582.]
CrossRef | Web of Science | Medline3
Cummings JH ,Roberfroid MB ,Andersson H , et al. A new look at dietary carbohydrate: chemistry, physiology and health. Eur J Clin Nutr 1997;51:417-423
CrossRef | Web of Science | Medline
To the Editor:
Fuchs et al. have failed to throw light on the cause of bowel cancer. They made no assessment of colonic function, so their article leaves unchallenged the evidence that bulkier stools and faster transit through the gut reduce the risk of bowel cancer.1 Nor does their study alter the fact that, for the person who wants bulkier stools and faster transit, increasing fiber intake — especially intake of wheat fiber — is usually effective.
1 ReferencesKenneth W. Heaton, M.D.
Stephen J. Lewis, M.D.
Bristol Royal Infirmary, Bristol BS2 8HW, United Kingdom1
Cummings JH ,Bingham SA ,Heaton KW ,Eastwood MA . Fecal weight, colon cancer risk, and dietary intake of nonstarch polysaccharides. Gastroenterology 1992;103:1783-1789
Web of Science | Medline
To the Editor:
One concern we have about the article by Fuchs et al. relates to the method used to measure dietary intake — that is, the reliance on a food-frequency questionnaire based entirely on self-reports and carried out on a large scale. The authors chose to report all data as adjusted for total energy intake, but there is no report of estimated total energy intake. We are concerned about the underreporting of food intake (a well-known bias among people who are aware of recommended healthy eating patterns).
Another source of concern is the fact that the average body-mass index reported is considerably lower than that of the American population at large. This discrepancy raises questions about the accuracy of the data collected with respect to the self-reports of weight.
Zecharia Madar, Ph.D.
Aliza Stark, Ph.D.
Hebrew University of Jerusalem, Rehovot, 76100 IsraelTo the Editor:
Fuchs and coworkers concluded that dietary fiber did not protect against colorectal cancer and adenomas. However, the data on fiber content from the 1980s do not reflect modern food-processing techniques. For example, extrusion cooking, which is used to process snacks and breakfast cereals, can depolymerize insoluble fiber molecules, forming smaller, more soluble fragments. Whether this processed soluble fiber has the same health benefits as naturally occurring soluble fiber (e.g., pectin) is debatable. Many of the health benefits attributed to dietary fiber are actually associated with fruits, vegetables, and grains rich in fiber. The walls of plant cells contain many phytochemicals in addition to fiber.
Spiller warned: “We must not be tempted to attribute effects of whole-fiber foods to the fiber fraction alone or to generalize the effect or lack of effect of an isolated polymer to the whole food when studying fiber or contradictory and misleading results will be the outcome.”1 Dietary fiber is believed to protect against cancer in a number of ways, one of which is the binding of dietary carcinogens. Although dietary fiber alone may not be responsible for all the health benefits attributed to it, researchers must take heed to avoid giving consumers the impression that fiber-rich foods are not healthful.
1 ReferencesMary Ellen Camire, Ph.D.
University of Maine, Orono, ME 04469Author/Editor Response
The authors reply:
To the Editor: Burkitt hypothesized that intake of dietary fiber accounted for the difference in rates of colorectal cancer between Africa and Western countries, but other factors could have been responsible. Consumption of red meat, alcohol, fruit, vegetables, and folic acid; cigarette smoking; body-mass index; and level of regular physical activity are all associated with the risk of colorectal cancer and are likely to explain these international differences.1
Ravin, Cummings and Southgate, Madar and Stark, and Camire question the validity of our measurements of fiber intake. The reproducibility and validity of our food-frequency questionnaires have been documented previously.2 The fact that, in our cohort and in others, strong inverse associations between fiber intake and symptomatic diverticular disease, coronary heart disease, hypertension, and non-insulin-dependent diabetes mellitus were observed in data obtained with the same questionnaire indicates that we used a physiologically relevant measure of fiber intake. Such benefits of dietary fiber were similarly observed in a parallel male cohort, although fiber intake was not associated with colorectal cancer or adenoma among these men.
Cummings and Southgate also question our adjustment of energy for dietary fiber. However, when dietary fiber was not adjusted for total energy, we found no relation between fiber intake and the risk of colorectal cancer.3
As Mohandas indicates, our study cannot rule out an effect of fiber consumed early in life on the risk of colorectal cancer. We did, however, assess the influence of fiber on the risk of the early precursor lesions of colorectal cancer, adenomatous polyps. During the 14-year follow-up period, we observed no reduction in the risk of colorectal adenoma with increasing intake of cereal, fruit, or vegetable fiber or total dietary fiber, even when we restricted our analysis to women who had undergone endoscopy before 1980 and were found to be free of polyps.3 Although we did not study an effect early in life, studies of migrants and trends over time within populations suggest that risk factors for colorectal cancer act in midlife or later.1
Madar and Stark raise the possibility of error arising from the use of self-reported weight. In a validation substudy of our cohort, self-reported weights were highly correlated with actual measurements (Spearman rank-correlation coefficient=0.96).4
As Camire indicates, there are cogent reasons for increasing fiber intake, particularly the inverse association observed with coronary heart disease in many studies.5 However, for the prevention of colorectal cancer, we would emphasize other strategies, for which the evidence is more convincing, for minimizing risk, such as avoiding red meat, alcohol, smoking, and obesity, as well as taking folic acid, participating in regular physical activity, and undergoing regular screening.
5 ReferencesCharles S. Fuchs, M.D.
Walter C. Willett, M.D.
Brigham and Women's Hospital, Boston, MA 021151
Giovannucci E ,Willett WC . Dietary factors and risk of colon cancer. Ann Med 1994;26:443-452
CrossRef | Web of Science | Medline2
Willett WC ,Sampson L ,Stampfer MJ , et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol 1985;122:51-65
Web of Science | Medline3
Fuchs CS ,Giovannucci EL ,Colditz GA , et al. Dietary fiber and the risk of colorectal cancer and adenoma in women. N Engl J Med 1999;340:169-176
Full Text | Web of Science | Medline4
Manson JE ,Colditz GA ,Stampfer MJ , et al. A prospective study of obesity and risk of coronary heart disease in women. N Engl J Med 1990;322:882-889
Full Text | Web of Science | Medline5
Willett W. Nutritional epidemiology. New York: Oxford University Press, 1998.
