1137 acids rather than neutral sterols as possible mediators of dietary effects on colon carcinogenesis. Other hypotheses have also attributed a role to bile acids, either directly as carcinogens or as precursors for bacterially-produced carcinogenic steroids ; 12 ring desaturation by certain clostridia could, it is speculated, yield carcinogenic aromatic compounds. 13 However, in case-control studies and in comparisons of high and low prevalence population samples, high faecal concentrations of both bile acids and sterols have been associated with colon cancer. 14-16 If bile acids are the fxcal steroids most likely, on present evidence, to have a promoting effect, several dietary components could influence this phase of tumorigenesis. The action of a high fat diet in increasing fxcal bile-acid excretion has been noted.’ Dietary fibre is in the limelight, partly on the basis of the epidemiological association of low intakes with a high prevalence of colon cancer.17.18 Dilution of bile acids and of carcinogens in a larger faecal bulk, shortened transit-times and sequestration of bile acids are among the proposed mechanisms for the apparent protective effect of fibre. Dietary cholesterol influences fwcal bile-acid excretion in the rat but only one of four studies’-6 indicates that this occurs in man. The comparison of a low-prevalence rural community in Finland with a high-risk New York population is of great interest.’6 Fat intake was similar in the two populations, fibre intake was greater in the former; although faecal bile-acid excretion per 24 h was similar in the two groups the concentration of faecal bile acids was far lower in Finland due to the three-fold greater sterol bulk. Neutral sterol excretion was in fact higher in the low risk
community. Hence the most strongly supported hypothesis at present is that a high-fat diet increases faecal bile-acid concentration which acts as a promoter of colonic cancer. Dietary recommendations to reduce faecal bile-acid concentration would include reduction of fat intake and an increase in those forms of fibre which increase sterol bulk. A restricted fat intake inevitably reduces dietary cholesterol too; in practical terms there may be little to choose between the fat and cholesterol hypotheses. Department of Chemical Pathology and Metabolic Disorders, St. Thomas’s Hospital Medical School, London SE1 7EH
introducing the suspected substance through a Ryle’s tube; for example, one patient said that tomato juice caused immediate pain but it did not happen when given this way. Another method is to
I have tested patients by without their knowledge
ask the dietitian
to
prepare food with and without the
sus-
pected ingredient though tasting the same. I have done this to show whether lactase deficiency was an incidental biochemical finding or causing the patient’s diarrhcea. The dietitian made up cakes, one batch with glucose and the other with lactose but with about the same sweetness; the patient took them home and reported any adverse reactions later. One hates depriving anyone unnecessarily of the pleasures of the table, and each of Dr Grant’s patients "reacted" to an average of ten different foods. Surely more work on this subject is needed before such an approach is universally adopted.
B
Queen Elizabeth Hospital, Birmingham B15 2TH
CLIFFORD HAWKINS
DELAY IN DIAGNOSIS OF COLORECTAL CANCER
SIR,-Mr Holliday and Professor Hardcastle (Feb. 10, p. 309) found a nine-month delay between the onset of symptoms and treatment in patients with rectal cancer in Nottingham. To correlate delayed hospital admission with the outcome of treatment, we reviewed 110 patients admitted to the surgical clinic of the Catholic University in Rome and operated upon for rectal cancer in the past ten years. (Details of methods will be published in the Italian Journal of Gastrcpnterology.) Very few patients had had sigmoidoscopy and double-contrast enemas before hospital admission; 11 % had liver metastases at operation; 41% of the surgical specimens showed lymph-node involvement (Dukes C); and radical surgery was possible in 76% of cases. Postoperative mortality (including
emergencies) was 7%. FREQUENCY
OF RISK FACTORS FOR AND COMPLICATIONS OF
SURGERY IN RELATION TO DELAY BETWEEN ONSET OF SYMPTOMS AND HOSPITAL ADMISSION
BARRY LEWIS
FOOD ALLERGIES AND MIGRAINE
SIR,-If Dr Grant (May 5, p. 966) hopes to convince readers that there is anything more in her treatment of migraine than faith healing, she must provide objective evidence. Elimination diets, used since the beginning of this century, have seldom provided evidence of true food allergy. For example, they were tried in the hope of finding a cure for ulcerative colitis, then discarded. The fact that symptoms such as a rise in pulse occurred when the suspected foods were re-introduced is no proof, especially since most of her patients already had functional symptoms (lethargy, depression, anxiety, flushing, dizziness, and so on). Patients with nervous dyspepsia become intolerant of various foods, which cause ill-effects, including vomiting; however, this "allergy" usually lies in the psyche and not in the gastric mucosa. It
is, of
course, easier said than done to
provide hard
data.
11. Reddy, B. S.,
Watanabe, K., Weisburger, J. H., Wynder, E. L. Cancer Res. 1977, 37, 3238. 12. Hill, M.J. C.R.C. crit.Rev. Toxicol. 1975, p. 31. 13. Hill, M.J.Am. J. clin.Nutr. 1974, 27, 1475. 14. Hill, M.J., Drasar, B.S., Aries, V.Lancet, 1970, i, 95. 15. Hill, M. J., Drasar, B. S., Williams, R. E. O., Meade, T. W., Cox, A. G., Simpson, J.E.P., Morson, B.G. ibid. 1975, i, 535. 16. Reddy, B. S., Hedges, A. R., Laakso, K., Wynder, E. L. Cancer, 1978, 42, 2832. 17. Burkitt, D. P. ibid. 1971, 28, 3. 18. Cummings, J. H., Hill, M. J., Jenkins, D. J. A., Pearson, J. R., AmJ.
clin. Nutr. 1976, 29, 1468.
When we looked at variables (e.g., anaemia or protein loss) which might influence the incidence of postoperative complications, we found that they did not vary with the interval between the onset of symptoms and hospital admission. The frequency of postoperative complications or death did not seem to be associated with the length of this interval (see table). Moreover, delayed referral to hospital did not seem to affect the possibility of radical surgery. This accords with Mr Irvin’s report (March 3, p. 489) that pathology of rectal cancer does not depend on duration of symptoms. Since a better prognosis can be achieved by diagnosis in the asymptomatic stage of the disease, an effort should be made to improve public knowledge and outpatient facilities. Routine mass screening for colorectal cancer would be very difficult in Italy, where there is no national programme for prevention or early diagnosis of digestive diseases, even though these constitute the most frequent cause of death from cancer.I M. PESCATORI Institute of Clinical Surgery, G. MINGRONE Catholic University, 00168 Rome, Italy G. MARIA
Wiggins, H. S. 1.
Ital. J Gastroent. 1978, 10, 67.