Time trends in colo-rectal cancer mortality in relation to food and alcohol consumption: United States, United Kingdom, Australia and New Zealand

Dietary fiber and colorectal neoplasia

PURPOSE

Dietary fiber has been implicated in colorectal neoplasia, despite conflicting evidence. This is a review of the currently available data on the role of dietary fiber in colorectal carcinogenesis.

METHODS

A literature search was conducted using the MEDLINE database. All case-control, longitudinal, and randomized, controlled studies published in English between 1988 and 2000 were identified, as were animal model studies in the period 1986 to 2000. Data from the various studies were tabulated and systematically analyzed, with particular emphasis on the effect of dietary fiber on tumor incidence and luminal parameters such as short chain fatty acids.

RESULTS

Epidemiologic correlation studies show a high intake of dietary fiber to be associated with a lower risk of colorectal neoplasia. Thirteen of the 24 case-control studies reviewed demonstrated a protective effect of dietary fiber against colorectal neoplasia, and 16 showed a protective effect of vegetables or vegetable fiber. On the other hand, of 13 longitudinal studies in various cohorts, only 3 demonstrated a protective effect of fiber and 4 a protective effect of vegetables or vegetable fiber. The five published randomized, controlled trials all investigated the effect of increased fiber intake on short-term adenoma recurrence; however, none showed any significant protective effect. Among 19 experimental studies in animal models, 15 showed a protective effect of fiber against tumor induction compared with controls. Animal studies also showed that poorly fermentable fibers (e.g., wheat bran and cellulose) were more protective than soluble fibers (e.g., guar gum and oat bran), which sometimes enhanced carcinogenesis. No clear correlation was found between luminal pH or short chain fatty acids and tumor induction.

CONCLUSIONS

On the basis of current data, there is little evidence to support the use of dietary fiber supplements to reduce the risk of colorectal neoplasia. Lifelong and early exposure may be important but are difficult to study. Other risk factors interact with the effects of dietary fiber.

Blood group antigen expression in the rat colon II. Modulation by dietary ethanol consumption

In the accompanying article, we established that in the rat distal colon expression of H, B, and Le(b) blood group antigens by goblet cells is phenotypically fetal in nature. Because of the cocarcinogenic property of ethanol, the present study examined the effects of dietary ethanol consumption, fasting, and withdrawal on the expression of these antigens in the adult rat colon. To that effect, male adult Sprague-Dawley rats were pair-fed ethanol-containing or control Lieber-DeCarli liquid diets for 3 weeks. The effects of ethanol withdrawal were studied in rats fed the ethanol-containing diet for 3 weeks followed by the control diet for 1, 3, and 6 days. In rats fed the control diet, no antigen expression in the distal colon was observed, as expected. Ethanol feeding for 3 weeks resulted in a striking reappearance of H, B, and Le(b) antigens in goblet cells of the distal colon. In colonic crypts, a lower-to-upper crypt gradient of increasing numbers of positive goblet cells was present, suggesting that the induction of antigen expression paralleled the differentiation of goblet cells. After an overnight fast, the number of positive cells was significantly decreased. Withdrawal of ethanol for 1 day further decreased the number of positive goblet cells. The decrease was reflected by a downward shift in the number of positive cells per crypt column, which was more striking in the lower and mid-crypt segments than in the upper segment, suggesting that antigen expression was more labile in immature differentiating goblet cells than in mature ones. No antigen staining of goblet cells was detected after 3 and 6 days of ethanol withdrawal. Hence, expression of H, B, and Le(b) antigens by goblet cells of the distal colon can be modulated by ethanol consumption. Expression in the distal colon of A and Le(a) antigens, which did not exhibit a fetal phenotype, was not affected by ethanol feeding. In conclusion, because of the oncofetal phenotype of H, B, and Le(b) antigens, their reappearance in the distal colon may serve as a cytochemical marker for early recognition of epithelial changes of the colon in ethanol-related cocarcinogenesis before more overt manifestations of neoplasia.

Dietary fiber and plant foods in relation to colorectal cancer mortality: the Seven Countries Study

Many observational studies have found that higher consumption of vegetables, and to a lesser extent of fruits, was associated with lower risk of colorectal cancer. In particular, fiber or foods high in fiber have received attention in the potential prevention of colorectal cancer. We performed an ecological analysis with data of the Seven Countries Study, to investigate whether intake of fiber and plant foods contributes to cross-cultural differences in 25-year colorectal-cancer mortality in men. In the Seven Countries Study, around 1960 12,763 men aged 40 to 59 were enrolled in 16 cohorts in 7 countries. Baseline dietary information was gathered in small random samples per cohort, and nutrient intakes were based on chemical analyses of the average diets per cohort. Crude and energy-adjusted mortality-rate ratios were calculated for a change of 10% of the mean intake of fiber and plant foods, i.e., total plant foods, fruits, vegetables, potatoes, grains, and related sub-groups. Fiber intake was inversely associated with colorectal-cancer mortality with an energy-adjusted rate ratio of 0.89 (95% confidence interval 0.80-0.97). An increase of 10 gram of daily intake of fiber was associated with a 33% lower 25-year colorectal-cancer mortality risk. Intakes of vitamin B6 [0.84 (0.71-0.99)] and alpha-tocopherol [0.94 (0.89-0.99)] were also inversely associated with risk. Consumption of plant foods and related sub-groups was not related to colorectal cancer. It appears that fiber intake best indicates the part of plant food consumption, including whole grains, that is relevant for lowering colorectal cancer risk.

Adherence to the European Code Against Cancer in relation to long-term cancer mortality: intercohort comparisons from the Seven Countries Study

Within the Seven Countries Study, we investigated whether population differences in 25-year cancer mortality and mortality due to cancer of the lung, stomach, and colorectum could be explained by population differences in adherence to the European Code Against Cancer. Baseline surveys were carried out around 1960 on 12,763 middle-aged men constituting 16 cohorts in seven countries; small samples of the cohorts kept dietary records. In 1987, food equivalent composites representing the average food intake of each cohort at baseline were collected locally and analyzed in one central laboratory. The vital status of all participants was verified after 25 years of follow-up. Overall adherence to the first four recommendations of the European Code Against Cancer was inversely related to 25-year total cancer mortality but not to all-cause mortality. The association with cancer mortality was essentially due to the inverse association for adherence to the guideline on smoking only. Each decrease in the percentage of smokers of 3.4% (10% of range) was associated with a relative risk of 0.93 of dying from cancer. Adherence to the recommendation on consumption of vegetables, fruits, and fiber-rich cereals was inversely related to stomach cancer mortality, whereas adherence to the recommendation on body weight, physical activity, and intake of fat was associated with higher stomach cancer mortality.

Alcohol and gastrointestinal cancer: pathogenic mechanisms

Chronic heavy alcohol consumption leads to a significantly increased risk of cancer in the oropharynx, larynx and the oesophagus. In the liver, chronic alcohol abuse results in cirrhosis, a precursor of hepatocellular cancer. More recentepidemiologic studies also demonstrate that regular alcohol consumption, even in low amounts, has an enhanced risk for rectal cancer and cancer of the breast. Alcohol by itself is not a carcinogen. However, alcohol can increase the susceptibility of various organs to chemical carcinogens by a variety of mechanisms. Among these, increased activation of procarcinogens through microsomal enzyme induction, a change in the metabolism and/or distribution of carcinogens, interference with the system that repairs carcinogen-induced DNA alkylations, direct mucosal tissue damage with consecutive stimulation of cellular regeneration and alcohol-mediated malnutrition may be of importance. In the upper gastrointestinal tract the production of acetaldehyde and free radicals via cytochrome P450 2E1 and via alcohol dehydrogenase may lead to tissue damage and to secondary hyper-regeneration. In addition, local mechanisms may also be involved in the co-carcinogenic process. In the rectal mucosa acetaldehyde seems to be an important factor in carcinogenesis and may be predominantly produced by faecal bacteria.

Role of bile acids in colorectal carcinogenesis

Dietary factors are considered important environmental risk determinants for colorectal cancer development. Epidemiological studies have shown that a high fat (or meat) intake is associated positively and a high starch, fibre (non-starch polysaccharide), vegetable and fruit intake negatively with colorectal cancer incidence. One mechanism by which these effects are possibly exerted is through the metabolism of secondary bile acids. Secondary bile acids are formed after enzymatic deconjugation and dehydroxylation of primary bile acids in the large bowel by anaerobic bacteria. It has been shown that these compounds can have tumour-promoting capacities in animal experiments. In epidemiological studies, colonic cancer risk is related to the faecal bile acid concentration. In serum and bile of patients with colonic adenomas, more deoxycholic acid was detected than in healthy controls. Secondary bile acids are toxic to several cell systems at physiological concentrations. The exact mechanism by which these amphiphilic molecules exert their action is not well understood. It might act through membrane damage, intracellular mitochondrial action or genotoxic effects. So far the evidence that bile acids are involved in colonic carcinogenesis is largely circumstantial. It is, however, well accepted that environmental factors, such as dietary habits influence genetic susceptibility. Bile acids could play a promoting role in this process.

Colorectal cancer risk factors: a case-control study in Bangkok

A case-control study for colorectal cancer risk factors was conducted in Bangkok, Thailand. A total of 279 incident cases of colorectal cancer were individually matched by sex, age and same hospital to 279 hospital controls with other cancers except gastrointestinal cancer. Each subject was interviewed with regard to bowel pattern information, family history, past history of illness and dietary information. The major findings were elevated risk for those with a history of bowel polyps (OR = 14.69, 95%CI = 2.01-301.46), parent's history of colon cancer (OR = 4.00, 95%CI = 1.39-12.43), anal abscess (OR = 3.78, 95%CI = 0.97-17.24), chronic colitis (OR = 3.61, 95%CI = 1.67-8.00), chronic hemorrhoid (OR = 3.13, 95%CI = 2.03-4.86) and the frequency of stools every three days or more (OR = 2.16, 95%CI = 1.17-4.01). The results also indicated an increased risk for dietary factors; bacon (OR = 12.49, 95%CI = 1.68-269.1) and butter (OR = 2.68, 95%CI = 1.29-5.68). There was a protective effect provided by banana (OR = 0.54, 95%CI = 0.37-0.79) and papaya (OR = 0.58, 95%CI = 0.40-0.84) for colorectal cancer. In unconditional logistic regression analysis, bacon showed the highest risk for colorectal cancer (OR = 8.82, 95%CI = 1.03-75.57), instead of bowel polyps (OR = 4.50, 95%CI = 0.48-42.59). The data suggest that nitrite-treated meat increases colorectal cancer risk while dietary fiber decreases colorectal cancer risk.

Moderate beer consumption and the risk of colorectal cancer

The relationship between beer consumption and the risk of colon and rectal cancer was considered in a case-control study conducted in northern Italy. The study was based on 828 histologically confirmed incident cases of colon cancer, 498 of rectal cancer, and 2,024 controls in hospital for a wide spectrum of acute, nonneoplastic, nonalcohol-related diseases. Beer drinking was reported by 6% of colon cancer cases, 7% of rectal cancer cases, and 10% of controls; regular beer drinkers (> or = 1 drinks/day) made up 2.6% of colon cancer cases, 3.2% of rectal cancer cases, and 4.1% of controls. Thus the multivariate relative risks (RR) for irregular drinkers were 0.6 [95% confidence interval (CI) 0.4-1.0] for colon and 0.7 (95% CI 0.4-1.2) for rectum. Corresponding values for regular drinkers were 0.7 (95% CI 0.4-1.2) for colon and 0.9 (95% CI 0.5-1.5) for rectal cancer. Despite the low frequency of beer drinking in this study, and hence its limited statistical power, the originality of the population in terms of colorectal cancer incidence, patterns of risk factor exposure, and the large dataset provide interesting and useful confirmation that moderate beer drinking is not associated with elevated colon or rectal cancer risk.

Reduced risk of colorectal cancer among recent generations in New Zealand

Male and female age standardised mortality and incidence rates of colorectal cancer have increased over the most recent 30 years in New Zealand. Among men and women aged 40 to 74, age standardised mortality and incidence rates increased 18 to 105%. However, age standardised mortality and incidence rates among younger men and women have declined from 14 to 69%. Analysis of trends in age specific mortality and incidence rates indicates that the occurrence of colorectal cancer has been declining equally for men and women in successive cohorts born about 1943 to 1953 in New Zealand. This decline in the frequency of colorectal cancer among recent generations was apparent for both the right and left sides of the colon and the rectum. Age-specific trends in coronary heart disease and breast cancer differed from those apparent for colorectal cancer, suggesting that the factors producing the reduction in colorectal cancer risk may affect these diseases among different age groups or may not be of major aetiological importance in these diseases. These trends provide empirical evidence that the occurrence of colorectal cancer can be reduced by at least 50% with a substantial component of the risk being determined before the age of 30. Further study is needed to establish whether changes in risk factors at older ages contribute to the prevention of the disease.

Cereal fiber, calcium, and colorectal cancer

Food intake during the preceding 15 years was evaluated in detail in 41 patients treated for colorectal cancer and an equal number of matched control subjects by means of a dietary history technique that permitted quantitation of nutrients. Dietary habits of the control group could be compared against two larger groups of 371 hospital control and 430 population control subjects. Patients with cancer, who were interviewed after complete recovery from surgery, consumed more fat, protein, and carbohydrates, and thus more energy, than control subjects although these differences were not statistically significant. Per unit energy, the habitual diet of patients with cancer contained less cereal fiber (P less than 0.001), less riboflavin (P less than 0.05), less calcium (P less than 0.05), and less phosphorus (P less than 0.05) than the diet of the control subjects. A high intake of either cereal fiber, total fiber, calcium, and phosphorus in relation to energy intake was found to be associated with a reduced risk ratio of colorectal cancer. For colon cancer separately, a high intake of calcium and cereal fiber was associated with a reduced risk ratio. For rectal cancer, a high intake of total fiber and cereal fiber was associated with a reduced risk ratio. High alcohol consumption correlated with an increased risk ratio. These data are compatible with previous Scandinavian studies relating food consumption to the incidence and mortality of colorectal cancer.

Tobacco smoking as a risk factor for colon polyps

BACKGROUND

Data from a cancer screening project among pattern makers were used to evaluate the association between tobacco smoking and prevalence of colon polyps.

METHODS

From 1981-1983, 549 White men were examined by flexible sigmoidoscopy and completed self-administered questionnaires including smoking histories.

RESULTS

One or more colon polyps were detected in 76 men. Standardized prevalence rates (SPR) for polyps increased by smoking category (never smoked = 0.094; ex-smokers = 0.118, current smokers = 0.214) and by cigarettes per day, years of smoking, and pack-years among both current and ex-smokers. Both adenomatous and hyperplastic polyps showed an association with smoking while other types of polyps and polyps with unspecified histology did not. The risk associated with smoking was greater for polyps greater than one centimeter in diameter. An interaction with occupational exposures was suggested by a greater increase in the SPR for polyps among current smokers employed as pattern makers for more than 10 years than among current smokers similarly employed for 10 years or less.

CONCLUSIONS

Since at least some colon polyps are considered precursor lesions to colon cancer, one of the most common cancers in the United States, this report suggests that the possible link between colon polyps and smoking deserves further evaluation.

Enhancing effect of ethanol or saké on methylazoxymethanol acetate-initiated large bowel carcinogenesis in ACI/N rats

Effects of ethanol or saké on intestinal carcinogenesis by methylazoxymethanol (MAM) acetate were examined in two experiments. In the first experiment, 39 male ACI/N rats were given two weekly intraperitoneal injections of MAM acetate (25 mg/kg body wt) and divided into two groups, with Group 1 being given 10% ethanol and Group 2 being given distilled water. The incidence of colonic cancer in Group 1 (15/17, 88%) was higher than in Group 2 (9/16, 56%, p = 0.040). Differences in the incidences of rectosigmoidal colonic neoplasms were even more marked (59% vs. 19%, p = 0.019) and their proportion of the total number of large intestinal neoplasms was greater in Group 1 (36%) than in Group 2 (15%, p = 0.046). In the second experiment, 97 female ACI/N rats were divided into 6 groups, with the animals of Groups 1-5 being given MAM acetate (2 times, 25 mg/kg body wt). Rats in Group 6 received saline. The rats received isocaloric drinks: Group 1, saké; Group 2, 50% saké; Groups 3 and 6, 15% ethanol; Group 4, 7.5% ethanol; and Group 5, nonalcoholic water. Incidences of rectosigmoidal colonic neoplasms in Groups 1, 2, and 3 (53%, 46%, and 50%) tended to be higher than in Group 5 (38%). Their proportions of the total number of large intestinal neoplasms in Groups 1 (68%) and 2 (67%) were slightly greater than in Group 5 (45%). The results suggest an enhancing effect of ethanol or saké on rectosigmoidal colonic carcinogenesis.

Cancer incidence in migrants to New South Wales from England, Wales, Scotland and Ireland

Cancer incidence in migrants to New South Wales (NSW) from individual countries within the British Isles has been compared with that in the Australian-born population using data from the NSW Central Cancer Registry for the period 1972-84. Indirectly age-standardised incidence ratios (SIR) showed that, for cancer at all sites combined, Scottish migrants had a significantly higher, and English migrants a lower, incidence than the native-born Australians. Melanoma of skin was less common in migrants from all four countries while lung cancer was more common. In all except the Irish migrants, stomach cancer was more frequent than in the Australian-born. Raised SIRs for bladder cancer were found in men from all the countries and for breast cancer in all except the Irish women but only in the English migrants were these ratios significant. English migrants differed from those from Wales, Scotland and Ireland in that, compared with the Australian-born, they had significantly lower SIRs for cancer of the colon (both sexes), head and neck, larynx and prostate (men), gallbladder and kidney (women), and a higher SIR for ovarian cancer. Bone cancer was relatively more common in men born in Wales. 'Other genital' cancers (penis and scrotum; vulva and vagina) tended to be more frequent in migrants from each country than in the Australian-born.

Colorectal cancer in the north and south of Ireland 1950-1984

STUDY OBJECTIVE

Northern Ireland has the highest standardised mortality ratios for colon cancer in the United Kingdom and the Republic of Ireland has some of the highest mortality rates for cancer in the world. The aim of the study therefore was to investigate trends in colorectal cancer in the north and south of Ireland over the period 1950 to 1984.

DESIGN

The study was a cohort analysis of deaths from colorectal cancer for ages 35-74 years by five year age groups, divided by sex.

SETTING

This was a population study involving all cases reported to the Registrar General of Northern Ireland and the Eire Vital Statistics and Central Statistical Office during the study period.

MEASUREMENTS AND MAIN RESULTS

As in mainland Britain, rectal cancer mortality declined in the north and the south during the study period, but the fall began sooner for males than females. Colon cancer mortality fell in the late 1950s but subsequently rose to its previous high levels.

CONCLUSIONS

The observation that there were declines in mortality in the north and south of Ireland in the late 1950s does not support the hypothesis that altered diet due to war rationing in Great Britain and Northern Ireland underlay the fall in British colon cancer mortality after the war. The very high standardised mortality ratios for colon cancer in Northern Ireland highlight a continuing major public health problem in the region.

Cancer incidence in migrants to New South Wales

Cancer incidence in migrants to New South Wales (NSW) from the British Isles, north-central, eastern and southern Europe, the Middle East and Asia has been compared with that in Australian-born residents using data from the NSW Central Cancer Registry for 1972-84. Indirectly standardized incidence ratios (SIRs) were low in migrants from all 6 regions for melanoma of skin and cancers of lip and, except in men from eastern Europe, colon. Oesophageal, rectal and prostatic cancers also tended to be relatively less common. Cancers which were more common than in the Australian-born were those of the stomach and, for men, bladder (except in the Asian-born). Migrants from different regions showed variations from the cancer pattern of the Australian-born population which, for the most part, were predictable from the known incidence of cancer in the countries within the region of origin. Exceptions were the high relative incidence of nasopharyngeal cancer in migrants from southern Europe and bladder cancer in men from all regions other than Asia.

Strategies for cancer prevention through diet modification

Diet and cancer research at the National Cancer Institute has grown from a budget of $2.5 million in 1974 to $55 million in 1988. The emphasis is partially on demonstrations of prevention strategies and chemoprevention trials. Studies to disseminate dietary goals in practical ways are undertaken with the aid of the food industry.

Types of dietary fat and the incidence of cancer at five sites

The specificity of a statistical association increases the likelihood that it represents a causal relationship. In exploring the relationship between dietary fat and cancer, specificity applies both to cancer sites (outcome) and to component fats (exposure). In this study, Armstrong-Doll criteria were used to select female cancer incidence data for breast, cervix, lung, and colon, and male incidence data for lung, colon, and prostate for 1973-1977 from 20 countries with reliable registry data. Truncated age-standardized rates were correlated with estimates of per capita disappearance of total fat and of saturated, monounsaturated, and polyunsaturated (total, fish omega-3, omega-6) fats in 1975-1977. Multiple regression analyses were standardized for estimated total calorie intakes and used to assess the association between each fat and incidence at each cancer site. Estimates of per capita dietary and crude fiber intakes were also included in the analysis. Total calorie intake was not associated with cancer at any site when controlled for total fat intake, whereas total fat intake was strongly associated with cancers of the breast, colon, and prostate even after adjustment for total calorie intake. Cancers of the lung and cervix were not correlated with dietary fat intake. Monounsaturated fat had no positive association with cancer at any site. Saturated fat was positively associated with incidence of cancers of the breast, colon, and prostate and polyunsaturated fat was associated with incidence of breast and prostate cancers but not colon cancer. Fiber intake, when included in the analysis, affected the magnitude of the fat-cancer correlations, particularly between total fat and colon cancer. Fish omega-3 polyunsaturated fat had a nonsignificant negative association with the cancer sites studied. The findings supported hypotheses based on the results of animal experiments showing that different kinds of fatty acids have different tumor-promoting capabilities.

Cancer incidence among Swedish brewery workers

In order to examine the risks of cancer, particularly of rectal cancer, among Swedish brewery workers, 6,230 men employed in the brewery industry in 1960 were followed-up during 1961-79 by the Swedish Cancer Registry. Using all Swedish men as a reference group, relative risks (RR) were computed with standardization for year of birth, year of follow-up, and geographic region. A total of 712 new cases of cancer were observed compared to 570.7 expected (p less than 0.001). Significantly increased risks were seen for several cancer sites, e.g. esophagus (RR = 2.5, 95% confidence interval (Cl) = 1.5-3.8), rectum (RR = 1.7, Cl = 1.3-2.3), pancreas (RR = 1.7, Cl = 1.2-2.3), and lung (RR = 1.4, Cl = 1.1-1.7). An excess risk of liver cancer was almost significant (p = 0.053, RR = 1.7, Cl = 1.0-2.8). The risk of colon cancer was not significantly increased (RR = 1.2, Cl = 0.9-1.5), and the difference between the relative risk of colon cancer and that of rectum cancer was nearly significant (p = 0.07). Our results support the hypothesis that high beer consumption is associated with an increased risk of rectal cancer.

Possible role of acetaldehyde in ethanol-related rectal cocarcinogenesis in the rat

Prospective epidemiologic studies have reported an increased risk of rectal cancer following chronic ethanol ingestion. The effect of ethanol on chemically induced colorectal carcinogenesis is controversial depending on the experimental conditions. In the present study the effect of chronic ethanol administration on acetoxymethylmethylnitrosamine-induced rectal cancer and the possible role of acetaldehyde in this process were investigated. Chronic ethanol administration resulted in an earlier occurrence of rectal tumors in this animal model. Because the concomitant administration of cyanamide, a potent acetaldehyde dehydrogenase inhibitor, showed a positive trend toward increased incidences of tumors, acetaldehyde could be involved in the ethanol-associated carcinogenesis. To measure colonic acetaldehyde, 12 chronically ethanol-fed and control rats received an acute dose of ethanol (2.5 g/kg body wt). The mucosal concentration of acetaldehyde was significantly higher in the rectum compared with the cecum (198 +/- 23 vs. 120 +/- 23 nmoles.g colon-1, p less than 0.05), but was not affected by chronic ethanol feeding. Furthermore, 6 germ-free rats had significantly lower acetaldehyde concentrations in the rectum (84 +/- 11 vs. 234 +/- 33 nmoles.g colon-1, p less than 0.01) and in the cecum (59 +/- 13 vs. 121 +/- 33 nmoles.g colon-1, p less than 0.05) compared with 6 conventional animals, and this was paralleled by the number of fecal bacteria in the 2 intestinal segments. In addition, to determine the effect of chronic ethanol feeding on colorectal cell turnover, 30 animals were pair-fed liquid diets. Using the metaphase-arrest technique, alcohol feeding induced rectal (19.1 +/- 2.0 vs. 9.1 +/- 1.8 cells.crypt-1.h-1, p less than 0.01), but not cecal (18.9 +/- 1.3 vs. 22.2 +/- 3.3 cells.crypt-1.h-1, p greater than 0.05) hyperregeneration. This was accompanied by an increase in the crypt proliferative compartment and increased mucosal ornithine decarboxylase activity (63 +/- 18 vs. 22 +/- 6 pmoles.hr-1.mg protein-1, p less than 0.05). The data show that chronic ethanol ingestion accelerates chemically induced rectal carcinogenesis and raise the possibility that acetaldehyde probably generated through bacterial ethanol oxidation may be involved in this process. The secondary hyperregeneration of the mucosa, observed after alcohol feeding, could by itself favour carcinogenesis.

A case-control study of male colorectal cancer in Aichi Prefecture, Japan: with special reference to occupational activity level, drinking habits and family history

The relationships of occupational activity level, drinking habits and family history of cancer to the risk of male colorectal cancer by subsites were investigated in a case-control study involving 1,716 cases with colon cancer, 1,611 cases with rectal cancer and 16,600 controls with other sites of cancer identified from the Aichi Cancer Registry, Japan 1979-1987. An occupation with a low activity level was associated with an increased risk of colorectal cancer; the age-adjusted relative risk (RR) compared to the high activity level group was 1.92 (95% confidence interval (CI): 1.38-2.67) for proximal colon cancer, 1.52 (95% CI: 1.19-1.94) for distal colon cancer and 1.38 (95% CI: 1.17-1.62) for rectal cancer. Beer drinkers showed an increased risk of colorectal cancer; the age-adjusted RR was 1.49 (95% CI: 1.13-1.95) for proximal colon cancer, 1.65 (95% CI: 1.34-2.04) for distal colon cancer and 1.88 (95% CI: 1.62-2.18) for rectal cancer. The RR for family history of colorectal cancer was 3.40 (95% CI: 2.19-5.29) for proximal colon cancer, 2.54 (95% CI: 1.73-3.75) for distal colon cancer and 1.78 (95% CI: 1.28-2.49) for rectal cancer. Multivariate analysis controlled for age, residence, marital status and smoking in addition to occupational activity level, beer drinking and family history of colorectal cancer did not materially change the RRs. When these three variables were combined, the RR was 15.72 (95% CI: 5.40-45.78) for proximal colon cancer, 10.55 (95% CI: 4.24-26.27) for distal colon cancer and 6.69 (95% CI: 3.12-14.36) for rectal cancer.

Cancer incidence in European migrants to New South Wales

The incidence of cancer in migrants to New South Wales (NSW) from Italy, Greece, Yugoslavia, Germany, the Netherlands, Poland and USSR has been compared with that in the Australian-born population using data from the NSW Central Cancer Registry for 1972-84. The indirectly age-standardized incidence ratios (SIR) in all seven countries were low for melanoma of skin and high for gastric cancer. Cancers of the colon, oesophagus and lip also tended to have low SIRs. Migrants from Italy, Greece and Yugoslavia had significantly less cancer at all sites than the native-born Australians mainly due to low SIRs for cancers of colon, lung (except Yugoslavian-born men), prostate and, in men, 'head and neck' (excluding nasopharynx). Cancers of breast and testis were relatively less common in migrants from Italy and Yugoslavia. SIRs were high for cancers of bladder (in Italian-born men), liver (in Greek- and Yugoslavian-born men) and nasopharynx (in Greek-born men and Italian-born men and women). Amongst migrants from the four more northerly European countries, ovarian cancer was relatively more common in women from Germany and Poland as was bladder cancer in men, but not women, from Germany and the Netherlands. Cancers which had significantly increased SIRs in one migrant group only were lung (Dutch-born men), cervix uteri and body of uterus (German-born women), gallbladder and bile ducts (Polish-born women), thyroid (Italian-born women), connective and other soft tissue (Russian-born men) and brain (Greek-born men and women computed together). Lymphomas were relatively less common in men born in Yugoslavia.

Hypotheses for the etiology of colorectal cancer--an overview

Colorectal cancer is the second most common cancer in the United States. Various dietary, colonic, and fecal components have been implicated as causative factors. Although numerous studies have been conducted to test them, so far no one factor has stood out as the most likely cause of colorectal cancer. This review presents the evidence for and against the major factors and concludes that bile acids are the most strongly implicated factors in the etiology of colorectal cancer.

Lifetime alcohol intake and risk of rectal cancer in western New York

The lifetime intake of total alcohol, beer, wine, and hard liquor was measured for 277 males and 145 females with pathologically confirmed, first, single, primary cancers of the rectum in western New York from 1978 to 1986. Controls who were age, sex, and neighborhood matched were also interviewed. Intake of beer and total alcohol was positively associated with rectal cancer risk. Most of the excess risk was found for the heaviest drinkers. Odds ratios for fourth quartile intakes for males were 1.80 (95% CI, 1.12, 2.89) for total alcohol and 1.86 (1.13, 3.06) for beer. No association was found with wine or hard liquor intake. Females drank considerably less in this population; trends were similar although not of as great magnitude as those for males. Adjustment for dietary risk factors did not change risk estimates appreciably. A high lifetime intake of beer and total alcohol was associated with an increased risk of rectal cancer, and this was independent of either socioeconomic status or diet.

The epidemiology and prevention of cancer in Australia

Based on the numbers of new cancers diagnosed, deaths from cancer and person-years of life lost as a result of death from cancer, non-melanocytic skin cancer, lung cancer, cancers of the colon and rectum, breast cancer, cancer of the prostate, melanoma, cancer of the stomach, cancer of the bladder, cancer of the pancreas, and non-Hodgkin's lymphoma were judged to be the most important cancers in Australia in 1982. Of these, only cancer of the pancreas appeared to be decreasing in frequency, and then only since 1979, while cancers of the lung and malignant melanoma of the skin were increasing rapidly. By the turn of the century the number of new cases of cancer diagnosed in Australia each year will be at least 50% higher than it was in 1982, mainly as a result of population growth and ageing. Given what is presently known about the causes of cancer it may be estimated that about one-third of cancers occurring in Australia could be prevented through feasible programmes. More than half of this change would be achieved through the elimination of tobacco smoking. A further 13% of cancer deaths, and some non-fatal cases, could be prevented by the effective implementation of cancer screening programmes of established efficacy. Even after this, however, many of the currently important cancers would remain significant. They are, therefore, priority subjects for research in cancer control.

Review of cancer among 4 religious sects: evidence that life-styles are distinctive sets of risk factors

The occurrence of various types of cancer have been reviewed and evaluated in 4 religious groups. These patterns have been critically assessed in light of the distinctive life-style features of these groups. All 4 religious groups considered in this paper have reduced overall rates of cancer, suggesting that the life-styles of all 4 groups have merit in terms of reducing the overall risk of cancer. The rate of smoking among these groups is nearly nil, and the lung cancer rate in all 4 of these religious groups is strikingly low. Cancer of the oral structures, pharynx, larynx, and esophagus is also generally quite low. Amish and Hutterites have unusually high rates of breast cancer and juvenile leukemia. Reproductive factors frequently mentioned as risk factors for breast cancer cannot explain the excess breast cancer in the Amish and Hutterite women because they should have had the effect of reducing the rate. None of the numerous risk factors, normally suggested for leukemia, are consistent with this observation. The observations on ovarian cancer tend to confirm low parity and late age at first birth as risk factors, although the evidence is not entirely consistent. Also, contrary to common observations, the pattern of ovarian cancer contrasts greatly with the breast cancer pattern, suggesting dissimilar risk factors. Their low rate of cervical cancer is consistent with promiscuity being a strong risk factor, but other frequently suggested risk factors were generally inconsistent with the observations. Cancers of the stomach, colon, rectum, urinary bladder and prostate, in these 4 religious groups, are not readily explained by the risk factors commonly implicated in cancer of these sites. The patterns of a few types of cancers were consistent with the prevailing opinions of risk factors, but some cancers were poorly explained and, in some cases, the cancer patterns contradicted commonly held opinions concerning risk factors. Religions that provide strong directives for the personal lives of adherents result in distinctive life-style, reflecting multiple disease related factors (risk factors and protective factors). Disease related factors are related to each other in simple or more complex ways (e.g. additive, multiplicative or even more complex). Therefore, when dealing with distinctive life-styles, it may be unwarranted to attempt to isolate individual risk factors.

Alcohol, physical activity and other risk factors for colorectal cancer: a prospective study

The aetiology of colorectal cancer was studied in a cohort of 11,888 residents of a retirement community. After four and one-half years of follow-up, 58 male and 68 female incident colorectal cancers were identified. Daily alcohol drinkers experienced nearly a two-fold increase in risk (2 sided P = 0.002). Colorectal cancer was also positively associated with Quetelet's index and inversely associated with avocational physical activity. The results were consistent for both sexes but were statistically significant only for males. With the exception of dietary vitamin C, none of the nutrients under study (i.e., vitamins A and E, dietary fibre, calcium, and beta carotene) showed a significant association with colorectal cancer. An inverse relationship between colorectal cancer and dietary vitamin C was observed in females, but there was no association with either vitamin C from supplements or with total vitamin C intake. Males and females who had 3 or more children showed a significantly reduced risk of colorectal cancer (RR = 0.45, 95% CI = 0.2, 0.9), but those with no children did not show the highest risk.

Colonic carcinogenesis: the microbial feast or famine mechanism

A mechanism is presented which suggests that high-fat, high-protein, low-fiber diets can cause an unfavorable microbial environment in the human colon which predisposes some individuals towards large bowel diseases. The digesta leaving the ileum on high-fat, high-protein, low-fiber diets has a high proportion of mucins, malabsorbed carbohydrates and proteins, bile acids, and sloughed epithelial cells. The irregular (pulsatile) emptying of rapidly fermentable ileal digesta into the colon causes a massive surge in microbial activity. The sudden availability of rapidly fermentable substrate generates a large microbial population in the exponential growth phase that soon depletes its substrate. For microorganisms to perpetuate until the next high-fat, high-protein, low-fiber meal propels ileal digesta into the colon, they must induce enzymes to ferment dying or dead microbes (cannibalism) in addition to colonic epithelial mucosa and mucins. As the carbohydrate-to-nitrogen ratio of the colonic contents decreases, the fermentation becomes more proteolytic and subacute levels of fermentation products such as ammonia may be generated. Carcinogens are concentrated within a small colonic mass and the probability of precancerous lesions and polyps developing in the colonic mucosa is directly related to the severity, frequency, and duration of these microbial "feast or famine" situations in the colon.

Case-control study of alcoholic beverages as etiological factors: the Melbourne Colorectal Cancer Study

As part of a large-scale investigation of colorectal cancer incidence, etiology, and survival, a case-control study was conducted to identify whether diet and alcohol, among other variables, were associated with the risk of colorectal cancer. This study compared 715 cases with 727 age- and sex-matched community controls. Findings from the dietary data are presented in the previous paper (Nutr Cancer 9, 21-42, 1987). The total life intake of specific alcoholic beverages was obtained from each subject. Data were classified by consumption of beer, wine, spirits, and alcohol. There was little evidence of an association of any of the alcohol variables with the risk of colon cancer. However, beer was found to be a risk factor for rectal cancer. This effect was more marked in males than in females, but the relative risks for females were consistent with those for males. Relative risk estimates changed only slightly when adjusted for the other alcohol variables and for the variables in the diet model; this suggests that the beer effect is separate from that of other alcohol variables and also from dietary variables. The age differences among beer consumers were found to be associated with cancer risk. Consumption of spirits was associated with a low risk for male rectal cancer. The risk of rectal cancer appeared to depend on beer drinking patterns in the previous 15-20 years.

Changing survival prospects in rectal carcinoma. A series of 1,306 patients managed by one surgeon

Cancer-specific survival prospects for rectal carcinoma in a series of 1306 patients managed from 1950 to 1979 by one surgeon worsened from 1970 to 1979. The prognosis was worse for all patients treated operatively from 1970 to 1979 compared with 1960 to 1969 (P less than 0.03). After potentially curative resection, survival was worse from 1970 to 1979 compared with 1950 to 1959 (P less than 0.02) and 1960 to 1969 (P less than 0.01), respectively; the corresponding five-year survivals were 72.5 percent, 72.3 percent, and 61.5 percent. The curative resection rate for the three decades was similar (66 to 70 percent). An increase in the incidence of Dukes' Stage C tumors from 23.3 percent to 32.3 percent (P less than 0.01) explains, at least partly, the decreased survival. The worsened survival prospects were not accounted for by changes in referral pattern, tumor site, or in the proportion of sphincter-saving resections performed. The worsening was paradoxically paralleled by earlier symptomatic presentation (P less than 0.001). Analyses of other Australian data are required to test the hypothesis that the worsened survival prospects are consequent to altered tumor biologic aggressiveness, possibly related to differences in the causal factors operating over the 30-year study period.

Epidemiology of polyps in the rectum and sigmoid colon. Evaluation of nutritional factors

Epidemiological studies have suggested an association between diet and colorectal cancer. Case/control studies, however, have been scarce, and studies based on interview with cancer patients who have symptoms from their cancer are inevitably prone to bias. An endoscopic population screening study for detection of colorectal adenomas enabled a double-blind registration of diet during 5 consecutive weekdays. Neither the participant nor the dietitian was informed of the findings at endoscopy. The estimation of 23 nutritional components was based on analysis of local commercial food and on the composition of foods in Norway. Results showed increasing consumption of fat and decreasing consumption of fiber and cruciferous vegetables in the presence of increasing neoplastic changes. The present material will form the basis for dietary-related follow-up studies.

Dietary factors in colon cancer: international relationships

The relationship between dietary factors and mortality from colon cancer was explored by an analysis of the correlation between age-adjusted colon cancer death rates for men in 38 countries and estimates of the availability of a number of dietary components. Cereals were the only source of fiber found to be negatively associated with colon cancer mortality after adjustment for the availability of total or animal fats, or total or red meats, foods that were themselves positively associated with mortality. The estimate of dietary fiber from cereals was more closely associated with mortality than that of crude fiber. The previously postulated protective effects of vitamins C and A and of cruciferous vegetables were not supported by the international data; we found no evidence of a negative association between colon cancer mortality and availability of these dietary factors. The positive association previously reported between colon cancer and beer consumption disappeared following adjustment for animal fat.

Dietary fiber and human cancer: critique of the literature

The relationship between dietary fiber consumption and risk of gastrointestinal cancer in humans is examined using representative studies of several types: international and intranational correlations, case-control analyses, metabolic investigations, cohort studies, and migrant studies. The strongest statistical association between diet and cancer is found in international studies in which numerous environmental variables differ. Studies on smaller groups within a single culture have not given strong support to the findings of international comparisons. Colon cancer rates within regions of the U.S. and other countries vary with sufficient magnitude that diet is unlikely to account for more than a minor proportion of risk. The evidence that a diet containing fiber-rich foods reduces risk of colon cancer must be considered tentative. Foods high in starch and fiber are statistically associated with a high rate of stomach cancer. Examination of the combined rates of colon and gastric cancer shows that the U.S. risk is low relative to countries in which a diet higher in fiber is consumed. It would be premature to suggest that a high fiber diet will confer protection against gastrointestinal cancer.

Dietary fibre consumption and its association with large bowel cancer in man

The hypothesis that lack of fibre in the diet is responsible for a variety of large bowel problems, including cancer, has stimulated much discussion and research over the past 15 years. However, the epidemiological examination of this hypothesis has been hampered by the absence of data on the fibre content of most of the world's foods. In studies in Britain and Scandinavia where consumption of the chemical fraction of dietary fibre, the non-starch polysaccharides, has been determined using accurate methods, significant negative association between colon cancer occurrence and NSP consumption have been shown. Fibre may therefore be protective to populations otherwise assumed to be at risk from a westernised type of diet. At present, methodological problems preclude the use of case-control studies in confirming or refuting these associations.

Fat intake and cancer of the gastrointestinal tract and prostate

The epidemiological evidence for and against the postulated role of dietary fat in the aetiology of human gastrointestinal and prostate cancers is critically reviewed. Evidence for a causal association is inconsistent, and much is of low validity. Several studies of colorectal cancer provide evidence to refute the fat hypothesis, at least in some populations. The expanding interest in precursor adenomas and the possible role of fat in their aetiology has led to case-control studies of colorectal adenomas and clinical preventive trials to see if fat is related to risk, but virtually no results have yet been published. The few studies done in relation to prostate cancer justify further investigation of the role of fat, but there appear to be no putative mechanisms to explain an association.

Descriptive epidemiology of colorectal cancer

The purpose of this report is to present the descriptive epidemiology of colorectal cancer using the most recent data available to highlight two characteristics of the disease. First is the great variation which takes place in the frequency of this disease over geographic areas of all sizes. Colorectal cancer is common in most countries of North America and Europe, is rare in Asia and is particularly uncommon in Africa. Internationally, the variation in colon cancer is 60-fold, and within Europe there is a 4-fold difference in the incidence of colon cancer between areas with the highest and lowest rates. For cancer of the rectum, variation internationally is 18-fold and within Europe it is 3-fold. Within the United Kingdom, colon cancer is uniformly higher in the 5 Scottish Cancer Registry Regions than in their counterparts in England and Wales. Even within Scotland there is a 4-fold range of colon cancer incidence in rates, with the North and South clearly demarcated by a striking difference in colon cancer incidence in both sexes. Secondly, examination of international mortality rates for colorectal cancer demonstrates remarkable differences in trends over time between countries. In countries where colorectal cancer mortality rates were initially low, rates have increased substantially. In many countries where rates circa 1950 were moderately high, they have increased slightly or become stabilized. However, in countries such as Scotland, Canada, England and Wales and the United States, where rates were initially high, there have been gradual falls in mortality over time.

Neither dietary ethanol nor beer augments experimental colon carcinogenesis in rats

There is an epidemiologic association between beer consumption and rectal cancer. Beer and ethanol were tested in the rat-dimethylhydrazine (DMH) experimental carcinogenesis model in order to verify this observation. Ethanol was found not to affect the number of colonic tumors induced by DMH (86 vs. 77 controls, P = 0.764). In rats fed beer and treated with DMH, there was a decrease in gastrointestinal tumor induction (P = 0.043). This instance then becomes one of many in which conclusions drawn from epidemiologic studies have been contradicted when subjected to analysis in an experimental colon carcinogenesis model.

Epidemiology of dietary fibre and colorectal cancer: current status of the hypothesis

The hypothesis that lack of dietary fibre in the diet is responsible for a variety of large bowel problems, including cancer, has stimulated much discussion and research over the past 15 years. However, the epidemiological examination of this hypothesis has been hampered by the absence of data on the fibre content of most of the world's foods. In Scandinavia and Britain where the consumption of the major chemical fraction of dietary fibre, the non-starch polysaccharides, has been measured using accurate methods, significant negative associations have been shown with colon cancer occurrence. These studies suggest that non-starch polysaccharides may be protective in populations at otherwise high risk of colon cancer from an excess of meat and fat. However, methodological problems in the assessment of non-starch polysaccharide consumption in individuals preclude the use of case control studies in verifying these associations within a single homogeneous population.

Is the changing pattern of colorectal cancer caused by selenium deficiency?

An hypothesis is presented to explain the changing pattern of colorectal cancer seen within the United States and other western countries in the last 30 years, the simultaneous increase of right-sided colonic cancer and disappearance of rectal cancer. Review of experimental and epidemiologic data suggests that this change may be due to a new systemic deficiency of the trace element, selenium. This deficiency has arisen not just from decreased consumption of selenium in the last 30 years, but also from increased consumption of zinc and fluoride, which may antagonize the effect of selenium.

Diet, alcohol, and relative weight in gall stone disease: a case-control study

A case control study of gall stone disease in relation to diet, alcohol, and relative weight was undertaken. The study population comprised 267 hospital patients with newly diagnosed gall stone disease, 241 individually matched controls selected from the community, and 359 controls who were patients in hospital. Dietary intake was estimated with a quantitative food frequency questionnaire. Multiple logistic regression analysis was used to estimate the net association between individual nutrients and the risk of formation of gall stones. Variations in risk with sex and age were examined in the light of prior evidence of influences of sex hormones and age on hepatobiliary metabolism. In both sexes increased intake of alcohol was associated with a decreased risk of developing gall stones; increased intake of simple sugars in drinks and sweets was associated with an increased risk; and increased intake of energy or fat was associated with an increased risk in young subjects. Obesity was associated with an increased risk only in young women.

Prospective study of alcohol consumption and cancer

The relation between alcohol consumption and the subsequent occurrence of the five most frequent cancers in Japanese men in Hawaii (cancer of the stomach, colon, rectum, lung, and prostate) was analyzed in a prospective study of 8006 subjects. Information on alcohol consumption was obtained through interviews in the mid-1960s, and the cohort has been followed since then. The analysis, which was adjusted for the effects of age and cigarette smoking, revealed a positive association between consumption of alcohol and rectal cancer, accounted for primarily by an increased risk in men whose usual monthly consumption of beer was 500 oz (15 liters) or more (relative risk, 3.05; P less than 0.01, as compared with those who did not drink beer). A significant positive relation between alcohol consumption and lung-cancer incidence was also found, accounted for primarily by an increased risk among subjects who consumed larger amounts of wine or whiskey, as compared with the risks among nonconsumers of these beverages (relative risk, 2.19, [P = 0.03] and 2.62 [P less than 0.01], respectively). No significant relation between alcohol consumption and the incidence of the other three cancers was found.

Alcohol, alcoholism, and cancer

Certain types of cancer are clearly associated with alcohol abuse, although the role of ethanol in carcinogenesis--as a carcinogen, co-carcinogen, promoter, or "innocent bystander"--is not known with certainty. The impact of alcohol abuse on the management of the patient with cancer is also discussed.