The search for the causes of breast and colon cancer

Cigarette smoking, alcohol, and the risk of colorectal adenomas

BACKGROUND

The present study was designed to further assess the reported association between cigarette smoking, alcohol, and colorectal adenomas.

METHODS

A number of environmental and life-style risk factors were examined in 236 patients with histologically proven adenomas and 409 controls with no adenomas.

RESULTS

Age, sex, race, and indication for procedure were similar in cases and controls. Those who had ever smoked were not at increased risk for adenomas compared with those who had never smoked. Years of smoking, cigarettes per day, and total pack-years showed no dose-response effect. Results for men and women were similar. Alcohol was a significant risk factor for men but not for women. Men in the highest quartile of daily caloric intake from alcohol were more than four times more likely than nondrinkers to develop adenomas, with a statistically significant trend in risk from the lowest to the highest quartile. These findings persisted after controlling for other potential risk factors for adenomas. The risk for colon and rectal polyps were similar. Men in the highest tertile of beer consumption were nearly six times more likely to develop adenomas than nondrinkers.

CONCLUSIONS

Beer drinking is a risk factor for colorectal adenomas in this population.

Colonic fermentation of complex carbohydrates in patients with familial adenomatous polyposis

Decreased production of butyric acid by colonic carbohydrate fermentation may predispose to colonic carcinogenesis, with the implicit assumption that the decrease in faecal butyrate found predates the development of the tumour. The influence of the genetic predisposition to colonic tumours and the presence of colonic polyps on in vitro fermentation of carbohydrates was examined. Stool samples from 11 normal controls and 20 patients with familial adenomatous polyposis (FAP) were incubated anaerobically with a range of carbohydrates. Fermentation patterns were similar for glucose and raffinose. These sugars produced different short chain fatty acid (SCFA) patterns from the two polysaccharides, starch and arabinogalactan, which differed one from the other. The FAP gene carriers with polyps produced less butyrate than normal controls (p < 0.005) and gene carriers without polyps (p < 0.05). There were corresponding decreases in the molar ratios of butyrate. Gene carriers without polyps produced less absolute amounts of acetate than normal controls (p < 0.05) and slightly less total SCFAs (p < 0.05) but were otherwise not significantly different. The decreased production of butyrate noted by other workers may be secondary to the tumours rather than a contributory cause.

Dietary fat promotes mammary tumorigenesis in MMTV/v-Ha-ras transgenic mice

The purpose of this study was to investigate the influence of dietary fat on mammary tumorigenesis in MMTV/v-Ha-ras transgenic mice. Female MMTV/v-Ha-ras transgenics were fed diets providing 0, 5 or 25% of calories from corn oil (CO). The mammary tumor incidence was 7% (0% CO), 36% (5% CO) and 52% (25% CO). Ras mRNA levels were increased in mammary tumors in the 25% CO group. The ras transgene was hypomethylated in mammary tumors, but not in liver or nontransformed mammary tissue. Mammary tumors expressed apolipoprotein E mRNA. Alterations in gene structure and expression in transgenic mice may suggest mechanisms by which dietary fat promotes mammary tumors.

Food-group consumption and colon cancer in the Adelaide Case-Control Study. I. Vegetables and fruit

Previous epidemiologic studies have shown an inverse association between vegetable and fruit consumption and risk of colon cancer. Vegetables and fruit contain a large number of potentially anti-carcinogenic substances, thus lending biological plausibility to this association. We conducted a case-control study in Australia, comparing 220 persons with histologically confirmed incident adenocarcinoma of the colon with 438 age- and gender-matched controls. Cases were identified via the South Australian Cancer Registry (1979-80); controls were randomly selected from the electoral roll. All participants completed a 141-item food-frequency questionnaire and were interviewed regarding demographic and other information. Consumption of 15 vegetable and fruit groups was investigated. Odds ratios (OR) for quartiles of consumption were derived using conditional logistic regression. All analyses were conducted separately for females and males. For females, greater intakes of onions and legumes were associated with decreased risk, with protein-adjusted OR of 0.48 and 0.53 respectively. Greater intakes of raw fruit and cabbage were associated with protein-adjusted OR of 0.76 and 0.71 respectively. For males, greater intakes of onions, green leafy vegetables, legumes, carrots and cabbage were associated with protein-adjusted OR in the range of 0.72 to 0.77. Consumption of potatoes was positively associated with risk in both genders. All 95% confidence intervals included 1.0. Analyses stratified by colon-cancer sub-site showed no strong and consistent differences between sub-sites for the vegetable and fruit associations. Results for meat, poultry, seafood, dairy foods and eggs are presented in a companion report.

Effects of dietary calcium and phosphate on the intestinal interactions between calcium, phosphate, fatty acids, and bile acids

Luminal free fatty acids and bile acids may damage the colonic epithelium and stimulate proliferation, which may increase the risk of colon cancer. It has been suggested that only soluble calcium ions (Ca2+) precipitate fatty acids and bile acids, thus reducing their lytic activity. Consequently, precipitation of luminal Ca2+ by dietary phosphate should inhibit these effects. To evaluate the proposed antagonistic effects of dietary calcium and phosphate, we studied the intestinal interactions between calcium, phosphate, fatty acids, and bile acids in rats fed purified diets that differed only in the concentrations of calcium and phosphate. Increased dietary calcium drastically decreased the solubility of fatty acids in the ileum, colon, and faeces, as well as the solubility of bile acids in the colon and faeces. Although dietary calcium strongly increased the total faecal fatty acid concentration and hardly affected the total faecal bile acid concentration, the fatty acid and bile acid concentrations in faecal water were drastically decreased by dietary calcium. Consequently, the lytic activity of faecal water was decreased. Dietary phosphate did not interfere with these intestinal effects of calcium. These results indicate that dietary phosphate does not inhibit the protective effects of dietary calcium on luminal solubility and the lytic activity of fatty and bile acids.

Tumour suppressor genes and colorectal neoplasia

Two distinct gene classes have been implicated in colorectal carcinogenesis. Tumour promoter genes (oncogenes, dominant oncogenes) produce an excessive positive stimulus to cell proliferation. The ras family of oncogenes are an example. Acquired mutations of the c-k-ras gene are commonly found in colonic adenomas and carcinomas. Tumour suppressor genes (anti-oncogenes, recessive oncogenes) normally constrain or regulate cell proliferation. Loss of this function through gene deletion or mutation is oncogenic. Inherited tumour suppressor gene mutations have now been identified in several of the familial cancer syndromes. Acquired tumour suppressor gene mutations are found in both sporadic and hereditary cancers. Together with the tumour promoter genes they provide the genetic basis for the cellular changes occurring during carcinogenesis. The retinoblastoma gene was the first human tumour suppressor gene to be characterized and exemplifies the class. More recently, linkage studies in the hereditary cancer syndromes and the detection of specific deletions in sporadic tumours have helped to identify several new tumour suppressor genes. At least four of these (MCC, APC, p53 and DCC) apparently contribute to sporadic colorectal carcinogenesis. Germ line APC mutations produce the inherited colorectal cancer syndrome familial adenomatous polyposis (FAP). Detection of these mutations using linked markers has already found clinical application in the screening of families with this disease. In the future, genetic diagnosis of hereditary non-polyposis colorectal cancer (HNPCC) and the recognition of those genetically susceptible to sporadic colorectal cancer may become possible. At the same time, as our understanding of the genes involved improves, new avenues for treatment and prevention of colorectal cancer may emerge.

Pathogenesis of colorectal cancer

This article takes the reader from colorectal lumen to the intracellular environment in an effort to explain the processes of tumorigenesis in colorectal cancer. The molecular changes seen in colorectal cancer development also are discussed, along with the potential for genetic manipulation and diagnostic molecular techniques. An era of major advances in understanding the pathogenesis of colorectal cancer has begun.

Dietary factors related to colorectal neoplasms

Although the etiology of colorectal cancer is multifactorial, experimental evidence suggests a role for dietary factors in the promotion of this disease. The complex interrelations governing energy balance and the consumption of fat, fiber, and micronutrients make it difficult to define the precise role of specific dietary factors in the etiology of colorectal neoplasms. Epidemiologic studies have demonstrated a correlation between the prevalence of colorectal cancer and per capita consumption of meat and fat. Case-control studies investigating the relation between colorectal cancer and dietary fat consumption have yielded inconsistent results. Prospective studies have failed to demonstrate a relation between fat consumption and subsequent risk for colorectal cancer. There is an inverse correlation between fiber intake and the prevalence of colorectal carcinoma. A more detailed analysis of the influence of various types of dietary fiber on the subsequent risk for colorectal cancer will provide a better understanding of this relation. Fiber derived from fruits and vegetables may provide more effective protection than cereal fibers. Correlational studies have established an association between total caloric intake and the prevalence of colorectal carcinoma. The design of future studies investigating the influence of individual dietary constituents on the risk for colorectal cancer must control for variations in energy balance as a confounding variable. Recent evidence suggests that a variety of micronutrients, including calcium, vitamin D, selenium, and vitamins A, C, and E, exert an anticarcinogenic effect. Studies designed to evaluate the influence of alcohol consumption on colorectal carcinogenesis have yielded inconclusive results. The potential influence of food preparation methods on colorectal carcinogenesis requires further evaluation. There is no conclusive evidence to support any of the hypotheses proposed to explain the role of dietary factors in colorectal carcinogenesis. Intervention trials designed to monitor intermediate markers for colorectal cancer such as increased epithelial-cell proliferation rates and the development of aberrant crypt architecture provide the opportunity for testing these hypotheses in relatively short-term studies. The results from such studies can be utilized in the design of large-scale, long-term prospective studies to evaluate the influence of dietary factors on the development of colorectal neoplasms. These trials should generate the information required to develop strategies for diet modification to reduce the incidence of colorectal carcinoma.

A sub-site-specific analysis of the relationship between colorectal cancer and parity in complete male and female Norwegian birth cohorts

Register and census data for complete cohorts of Norwegian men and women born between 1935 and 1969 have been used to examine the relationship between reproductive factors and the incidence of colorectal cancer. Among 1.1 million men and 1.1 million women under observation, 491 male and 859 female cases of colorectal cancer were diagnosed during the period of follow-up. Our hazard model estimates clearly show that in these young cohorts, women with 2 or more children run a lower risk of having a malignant tumor in the cecum or ascending colon than do other women of the same age and in the same birth cohort. No association with parity is found with respect to cancer in the transverse or descending colon, whereas a downward trend in the effect estimates, followed by an upturn, appears for the rectum, sigmoid colon and rectosigmoid junction. Such correlations, which also hold when we control for education and place of residence, are not found for men. The observed relationship between parity and cancer incidence is not likely to be exclusively explained by life-style differentials. Presumably, there is a biological effect of the number of pregnancies and deliveries, net of age at first birth and other reproductive factors, on the development of colorectal cancer.

Chromosome analysis of 20 breast carcinomas: cytogenetic multiclonality and karyotypic-pathologic correlations

Short-term cultures from 20 breast carcinomas were analyzed cytogenetically. A normal female chromosome complement was found in 4 cases. Clonal chromosome aberrations were detected in 16 tumors. In 10 tumors, multiple cytogenetic clones were found; in 2 cancers the clones were related, reflecting clonal evolution, but in the remaining 8 tumors the clones were cytogenetically unrelated, indicating clonal heterogeneity in the origin of the tumor parenchyma. Correlation analysis between karyotypic and pathologic parameters indicated that cases with complex karyotypes and/or cytogenetically unrelated clones, when compared with cases with a single simple karyotypic abnormality, were generally of higher histologic malignancy grade, had more mitoses in the histologic sections, and also more often had carcinoma in situ lesions in the same breast.

Mitotic activity in colorectal mucosa of healthy subjects in two Italian areas with different dietary habits

The proliferative activity was evaluated in colorectal biopsies of 39 healthy subjects living in two distinct geographical areas, Trieste in northern and Florence in central Italy. Subjects living in Trieste had a significantly higher mitotic activity compared with subjects living in Florence (mitoses/cells counted x 100 were 0.17 +/- 0.04 in Trieste and 0.089 +/- 0.02 in Florence). The results of a dietary questionnaire also showed that subjects in Trieste consumed significantly fewer starches, fibers, nitrites, and proteins. However, no correlation was evident between the consumption of these nutrients and intestinal proliferation.

Dietary factors and breast cancer risk in Vaud, Switzerland

The relationship between dietary factors and the risk of breast cancer was investigated in a case-control study conducted in the Canton of Vaud, Switzerland as a pilot phase for a larger cooperative study within the SEARCH Programme of the International Agency for Research on Cancer (Lyon, France). A total of 107 incident, histologically confirmed cases of breast cancer and 318 controls admitted to hospital for acute, nonhormone-related, gynecological, metabolic, or neoplastic disorders were interviewed. Significant direct trends in risk were observed with total energy intake [relative risk (RR) for highest vs. lowest intake tertile = 1.9] and, after allowance for energy intake, with frequency of consumption of various types of meat (RR = 2.1 for the highest tertile), cheese (RR = 2.7), and alcohol (RR = 2.1). Significant protections, on the order of 40-60% reductions for the highest vs. lowest consumption tertile, were conferred by total green vegetable consumption, selected types of vegetables and fruits (cucumbers, onions, pears), and a summary index of beta-carotene intake (RR = 0.4 for highest consumption tertile). Thus the present study confirmed the existence of an unfavorable dietary pattern for breast cancer risk (characterized by high-calorie, selected sources of animal fat and alcohol intake). Moreover, a significant protection could be gained by consuming a diet rich in vegetables and perhaps fruits.

Colorectal adenomas and diet: a case-control study of subjects participating in the Nottingham faecal occult blood screening programme

Diets high in animal fat and protein and low in fibre and calcium are thought to be factors in the etiology of colorectal cancer. Intakes of these nutrients were determined in three groups participating in a randomised trial of faecal occult blood (FOB) screening. A diet history was obtained by interview from 147 patients with colorectal adenomas, 153 age and sex matched FOB-negative controls (a) and 176 FOB-positive controls without colorectal neoplasia (b). Unconditional logistic regression was used to estimate relative risks (RR) and 95% confidence limits (increases) adjusted for age, sex and social class. After adjustment for total energy intake, no associations were found with total, saturated or mono-unsaturated fat, or calcium intake. For total fibre intake there were non-linear relationships with both control groups with the crude RR for highest quintiles of total fibre intake compared to the lowest being 0.6, although this pattern was no longer apparent after adjustment for energy intake with group (a). In comparison with group (b) cereal fibre intake showed a more consistent inverse relationship with adenoma prevalence with the RR for ascending quintiles of intake being 1.0, 0.7 (0.3-1.6), 0.5 (0.3-1.1), 0.7 (0.4-1.4) and 0.3 (0.1-0.6) (trend chi 2 = 8.80, p = 0.003). In comparison with group (a), the adjusted RR for the highest quintile of cereal fibre intake compared with the lowest was 0.6, but no clear trend was apparent. There was an unexpected positive relationship between adenomas and polyunsaturated fat intake with the RR for having an adenoma being 1.0, 2.8 (1.3-6.1), 1.6 (0.7-3.4), 3.5 (1.6-7.5) and 2.3 (1.1-5.0) for ascending quintiles of polyunsaturated fat intakes (trend chi 2 = 4.8, P = 0.03) in comparison with group (a) only. Our data, while providing no support for the role of dietary animal fat or protein, do support the protective role of dietary cereal fibre in the etiology of colorectal adenomas.

The growth of preneoplastic lesions by 1,2-dimethylhydrazine in rat colon is inhibited by dietary starch

The effect of dietary starch and sucrose on colon proliferation and the growth of foci of dysplastic crypts in the colon (FDC) were studied in female Sprague-Dawley rats, treated p.o. with 1,2-dimethylhydrazine (DMH). The animals were fed for 30 and 105 days with high fat (23% w/w corn oil) diets in which carbohydrates were represented by corn starch (starch diet) or sucrose (sucrose diet) (46% w/w). After 105 days of feeding, proliferation was markedly reduced in animals fed the starch diet. The number of FDC was not significantly affected by dietary treatments. However, after 30 and 105 days the percent of small FDC (formed by one-two dysplastic crypts) was higher in the animals fed the starch diet when compared to the sucrose diet. In the cecum of the animals fed the starch diet the percent of butyrate, propionic, isovaleric, and valeric over total short-chain fatty acids (SCFA) was increased, whereas the percent of acetic acid was decreased. Cecal pH was also decreased in the animals fed starch. The results suggest that starch diets have a protective role against DMH-colon carcinogenesis in the rat, mediated by a drop in cecal pH and an increased concentration of some SCFA.

Relationship of diet to small and large adenomas of the sigmoid colon

The relation of dietary factors to the risk of adenomas of the sigmoid colon was examined in men receiving a retirement health examination at the Self-Defense Forces Fukuoka Hospital between October 1986 and 1990. A total of 187 adenoma cases and 1557 controls with normal colonoscopy were identified in the series. Cases were further classified into small-adenoma (< 5 mm, n = 78) and large-adenoma (> or = 5 mm, n = 67) groups. The consumptions of selected foods and beverages were ascertained before colonoscopy by means of a self-administered questionnaire. After adjustment for smoking, alcohol use, rank and body mass index, low rice consumption and high meat intake were independently associated with an increased risk of large adenomas. The risk of small adenomas was not related to either rice consumption or meat intake. Adjusted odds ratios of large adenomas for the low, intermediate and high consumption levels of rice were estimated to be 1.0 (referent), 0.83 and 0.43, respectively (trend P = 0.08), and the corresponding figures for meat consumption were 1.0 (referent), 1.58 and 2.38, respectively (trend P = 0.02). The findings suggest that low rice consumption and high meat intake may promote the growth of colon adenomas, thereby increasing the risk of colon cancer.

Medical genetics--genes and GPs for the 1990s

DNA manipulation techniques have allowed the isolation and identification of many genes which cause disease in man. The same techniques are now being automated, and are on the threshold of application to more common and more complex genetic predispositions to diseases such as diabetes and schizophrenia. This article outlines the technology behind this development, and discusses some of the social, legal and ethical difficulties that widespread screening for genetic traits will raise. It is argued that these issues must be considered now, so that considered social responses to this technology are in place in the five years it is likely to take to develop. The techniques of DNA manipulation that have led to recombinant proteins and biopharmaceuticals have also led to an unprecedented increase in understanding of human genetics, and a corresponding increase in the practical side of human genetics--medical genetic diagnosis. As a result, medical genetics is moving from a specialty of statisticians and cytologists into the mainstream of medicine, and will soon be the concern of every hospital, and probably of every GP. This offers huge benefits to patients, but also threatens misunderstanding and even the potential of abuse if this sophisticated understanding of our fundamental natures is used without precision.

Familial risk of colo-rectal cancer in a low incidence area in southern Italy

The risk of colo-rectal cancer (CRC) in subjects with a positive family history (FH+) for malignancy has been assessed by means of a case-control study carried out between 1987-89 in an area of about 215,000 inhabitants in Southern Italy. One hundred and nineteen CRC cases were compared with 119 sex- and age-frequency matched population controls. Detailed pedigrees were collected at the family homes of both cases and controls. The odds ratio (OR) of CRC, adjusted by means of logistic regression for age, sex and number of first-degree relatives, increased with the number of any kinds of cancers in first-degree relatives with a significant linear trend (p = 0.042), while there was no risk with a FH+ for digestive cancer excluding CRC or for other cancers excluding large bowel and digestive organs. The OR (and 95% confidence interval) for CRC was 5.9 (1.64-21.23) for at least one first-degree relative with CRC. After a mutual adjustment between CRC and the other cancers in the families of cases and controls, the risk of CRC with a FH+ for other cancers did not change, revealing a strong association (p = 0.002) for CRC alone. From the analysis of the family history of cancer in the case group, the relative frequency of families that satisfied the criteria for so-called hereditary non-polyposic colo-rectal cancer (HNPCC) was 2.6%. The increased relative risk of CRC observed only in families with FH+ for CRC is a supportive finding for organizing and planning prevention and genetic counselling for these families, whose members should be referred for further assessment.

Diet and colorectal cancer: results of two case-control studies in Russia

Two case-control studies of diet and colorectal cancer were conducted in Moscow and Khabarovsk. The Moscow study comprised 100 cases of colorectal cancer and 100 neighbourhood controls. The Khabarovsk study consisted of 117 cases of colorectal cancer and 117 population controls. A history of the usual dietary intake one year prior to interview was taken using a food frequency questionnaire. Effects were adjusted in analysis for energy intake and education. Significantly reduced risks were observed with high intakes of cellulose (P = 0.001), beta-carotene (P = 0.002), vitamin C (P = 0.007), polyunsaturated fatty acids (PUFA) (P = 0.004), cholesterol (P = 0.04), and with a high ratio of PUFA to saturated fatty acids (SFA) (P = 0.002). Significant increases in risk were observed in association with high ratios of intakes of protein to cellulose (P = 0.002) and of fat to cellulose (P = 0.008). High intake of total fat was associated with non-significant decrease in the risk (P = 0.12), while high intake of SFA resulted in statistically non-significant increase in risk (P = 0.40). Significant reductions in risk were associated with high frequencies of consumption of vegetables (P = 0.001) and fruit (P = 0.009). There were results suggestive of a decreased risk with a high frequency of milk consumption (P = 0.06) and an increased risk in association with a ratio of meat to vegetable frequencies (P = 0.09). After adjustment among factors effecting risk of colorectal cancer statistically significant increase in the risk was seen only for protein/cellulose ratio and significant protective effect for PUFA/SFA ratio, beta-carotene and vegetable consumption.

Colorectal adenomas and energy intake, body size and physical activity: a case-control study of subjects participating in the Nottingham faecal occult blood screening programme

Most case-control studies of colorectal cancer have shown a positive association with energy intake. In contrast studies which have considered physical activity have found the most active to have a lower risk of colonic cancer and obesity appears to be no more than weakly related to colorectal cancer. We therefore compared energy intake determined by a diet history interview, self-reported height and weight, together with measures of lifetime job activity levels and leisure activity in the year prior to interview in 147 cases with colorectal adenomas and two control groups (a) 153 age-sex matched FOB-negative subjects (b) 176 FOB-positive subjects in whom no adenoma or carcinoma was found. Unconditional logistic regression was used to estimate relative risks (RR) and 95% confidence intervals () adjusted for age, sex and social class. No association with weight or body mass index was found. The only association with physical activity found with both control groups was an inverse association with running or cycling for half an hour continuously at least once a week RR 0.46 (0.2-1.3) compared with control group (a), and RR = 0.32 (0.1-0.8) compared with (b), but few subjects engaged in such activity. There was an inverse association with energy intake (trend chi 2 = 5.3, P < 0.025) in the comparison with control group (a) only, a finding which is consistent with those of two previous studies of asymptomatic adenoma.

Mandibular osteomas in sporadic colorectal carcinoma. A genetic marker

Pantomography of the mandible was performed in 98 patients with sporadic colorectal adenocarcinoma. Twenty-eight patients (29%) had osteomas versus 5% in a control group (P < 0.001). Mandibular osteomas are found in most patients with the premalignant dominant syndrome familial adenomatous polyposis. Sporadic colorectal cancer examinations of married couples have shown that diet has only a moderate influence on the development of colorectal cancer, whereas pedigree studies indicate a genetic component. On this basis we conclude that mandibular osteomas are probably genetic markers of the development of sporadic colorectal carcinoma.

The effects of soluble-fiber polysaccharides on the adsorption of a hydrophobic carcinogen to an insoluble dietary fiber

Dietary fiber is believed to decrease the incidence of colorectal cancer, but not all types of fiber are equally protective. Dietary fibers may be divided broadly into insoluble and soluble fibers, and there is evidence from animal experiments that the latter not only fails to protect against colorectal cancer but may enhance its development. Adsorption of carcinogens to insoluble dietary fiber in the intestinal tract is one of the mechanisms by which dietary fiber is believed to protect against colorectal cancer. In previous in vitro experiments, we showed that the hydrophobic carcinogen 1,8-dinitropyrene (DNP) adsorbs to insoluble plant cell wall components (insoluble dietary fibers). Soluble polysaccharides (pectic polysaccharides) extracted from the walls of parenchyma cells of dicotyledonous plants were found to maintain DNP in aqueous solutions and decrease its adsorption to insoluble wall components. In the present study, we examined a commercial preparation of pectin and seven other soluble-fiber polysaccharides with diverse structures for their effects on the distribution of DNP. Many of these are used as emulsifiers and stabilizers in the food industry. They all maintained DNP in aqueous solution and decreased its adsorption to alpha-cellulose, which we used as an example of an insoluble dietary fiber. Gum arabic was the most effective and kappa-carrageenan the least. The capacity of the polysaccharides to act as emulsifiers and stabilizers may explain their effects on DNP distribution. The monosaccharide glucose and the disaccharide cellobiose had no effect on the distribution of DNP. These results indicate three possible mechanisms by which soluble-fiber polysaccharides may enhance the development of colorectal cancer. First, because they reduce the ability of insoluble dietary fibers to adsorb hydrophobic carcinogens, more carcinogens may enter the colon maintained in solution than adsorbed onto insoluble fibers. Second, if soluble-fiber polysaccharides are maintaining hydrophobic carcinogens in solution and these polysaccharides are degraded by bacterial enzymes in the colon, then the carcinogens may come out of solution and be deposited onto the mucosal surface of the colon. Third, soluble-fiber polysaccharides may cross the intestinal epithelium and carry with them carcinogens maintained in solution. These studies have important consequences for nutrition, because soluble-fiber polysaccharides represent a common component of foods.

Long-term morbidity and mortality of overweight adolescents. A follow-up of the Harvard Growth Study of 1922 to 1935

BACKGROUND

Overweight in adults is associated with increased morbidity and mortality. In contrast, the long-term effect of overweight in adolescence on morbidity and mortality is not known.

METHODS

We studied the relation between overweight and morbidity and mortality in 508 lean or overweight adolescents 13 to 18 years old who participated in the Harvard Growth Study of 1922 to 1935. Overweight adolescents were defined as those with a body-mass index that on two occasions was greater than the 75th percentile in subjects of the same age and sex in a large national survey. Lean adolescents were defined as those with a body-mass index between the 25th and 50th percentiles. Subjects who were still alive were interviewed in 1988 to obtain information about their medical history, weight, functional capacity, and other risk factors. For those who had died, information on the cause of death was obtained from death certificates.

RESULTS

Overweight in adolescent subjects was associated with an increased risk of mortality from all causes and disease-specific mortality among men, but not among women. The relative risks among men were 1.8 (95 percent confidence interval, 1.2 to 2.7; P = 0.004) for mortality from all causes and 2.3 (95 percent confidence interval, 1.4 to 4.1; P = 0.002) for mortality from coronary heart disease. The risk of morbidity from coronary heart disease and atherosclerosis was increased among men and women who had been overweight in adolescence. The risk of colorectal cancer and gout was increased among men and the risk of arthritis was increased among women who had been overweight in adolescence. Overweight in adolescence was a more powerful predictor of these risks than overweight in adulthood.

CONCLUSIONS

Overweight in adolescence predicted a broad range of adverse health effects that were independent of adult weight after 55 years of follow-up.

Evidence of prenatal influences on breast cancer risk

Intrauterine exposure to high concentrations of endogenous pregnancy oestrogens may be important in the aetiology of breast cancer. In a nested case-control study we have assessed the relation between breast cancer risk and indicators of pregnancy oestrogen concentrations; pre-eclampsia/eclampsia is negatively related and measures of fetal size are positively related to oestrogen concentrations. Standard records for women born at Uppsala University Hospital between 1874 and 1954 were linked with records of invasive breast cancer cases, identified through their unique national registration numbers in the Swedish Cancer Registry during 1958-90. For each breast cancer case, we selected as potential controls female offspring of the first three mothers admitted to the hospital after the case's mother; only controls still living in Sweden and free from breast cancer when it was diagnosed in the case were finally included. Conditional logistic regression analysis was done for 458 breast cancer cases and 1197 matched controls. Pre-eclampsia/eclampsia was associated with a breast cancer rate ratio of 0.24 (95% confidence interval 0.09-0.70, p = 0.01). Linear trends for breast cancer incidence with increasing birth weight, birth length, and placental weight were positive but not significant. Thus, prenatal factors are important in breast carcinogenesis. Concentrations of pregnancy oestrogens may be one such factor, but other prenatal or perinatal factors cannot be excluded.

Dietary fat and breast cancer

High-fat diets are thought to increase the risk of breast cancer because animals develop mammary cancer more readily when they are fed high-fat compared to low-fat diets, and breast cancer incidence and mortality are higher in countries with high-fat as compared to those with low-fat diets. Prospective cohort studies and case-control studies have failed to provide much support for this theory, but such studies are less capable of showing the relationship because of smaller differences in dietary fat intakes of the study populations; difficulties in assessing the diets of individuals over a period of time; and possible differences in genetic susceptibility of cases and controls to breast cancer. Studies on migrants have shown that breast cancer incidence and mortality increase in populations who move from countries with low-fat to those with high-fat diets, indicating that observed geographical differences in breast cancer are due to environmental rather than genetic factors. This is supported by time-trend studies showing that breast cancer increases in countries as the level of fat in the diet rises. Controlled, long-term dietary trials are needed to determine whether the converse is true: namely, that reduction of dietary fat can reduce the risk of breast cancer. Large groups are required to achieve statistical significance, but smaller numbers may be adequate for studies on high-risk individuals. Preliminary experiments already have demonstrated the feasibility of carrying out such dietary trials.

Dietary fat and breast cancer risk: the feasibility of a clinical trial of breast cancer prevention

Animal experimental evidence and human ecological data suggest that dietary fat intake is related to breast cancer risk. Epidemiological studies within countries have given inconsistent results but are limited by the restricted range of dietary intake found in Western populations and by error in the measurement of fat consumption. Experimental evidence, derived from controlled clinical trials in which the range of fat intake is increased beyond that seen in most Western populations, is capable of overcoming this limitation of observational epidemiology, and would provide the strongest evidence available concerning the relationship of dietary fat intake to breast cancer risk. Further, such trials are the only means likely to answer the question of whether breast cancer risk in high-risk subjects can be modified by changing dietary fat intake. We describe here several aspects of the feasibility of an experimental approach to this problem, including the identification of subjects at increased risk for breast cancer, and the demonstration that such subjects will enter a clinical trial of dietary fat reduction and comply with a low-fat diet. It is shown that subjects can be recruited and retained in such trials, that satisfactory dietary compliance can be achieved over at least 24 mon and that the subjects selected are at demonstrably increased risk of breast cancer. This finding indicates that it is feasible to test the dietary fat-breast cancer hypothesis experimentally by means of a clinical trial.

Primary prevention of colorectal cancer through dietary modification

Because colorectal cancer is the second leading cause of cancer death in the United States and many developed countries, its primary prevention is of extreme importance. Environmental and dietary factors are considered responsible for 85-90% of all cases. Epidemiologic, animal, and biochemical studies suggest that diets high in total calories and fat and low in various dietary fibers, vegetables, and micronutrients are associated with an increased incidence. Of these factors, calcium and wheat bran have been used most extensively in recent trials. It has been reported that 1.5-2.0 g/day of calcium significantly decreases DNA synthesizing cells of high-risk patients. However, chronic wheat bran supplementation appears to decrease both rectal mucosal DNA synthesis and polyp recurrence. Several clinical trials currently are underway to evaluate the diet-colon cancer link. The results of these studies will help to determine the importance of dietary intervention in the reduction of the colorectal cancer risk.

Serum cholesterol level, body mass index, and the risk of colon cancer. The Framingham Study

BACKGROUND

Some studies have linked low serum cholesterol levels to increased risk of colon cancer, particularly in men. Results have been inconsistent, with preclinical disease frequently offered to explain any apparent association.

METHODS

The Framingham Study cohort of 5209 persons, initially 30-62 years of age and observed more than 30 years, was evaluated. Baseline data included lipoprotein fractions, total cholesterol levels, body mass index, alcohol intake, and cardiovascular risk variables such as cigarette smoking, hypertension, and glucose intolerance.

RESULTS

In this population, colon cancer in men is related inversely to serum cholesterol levels, even when the first 10 years of follow-up are eliminated to reduce the effect of preclinical disease. This effect is concentrated in the Svedberg 0-20 fraction, corresponding to low-density lipoprotein levels. Another finding only in men is the direct relation of body mass index to colon cancer incidence.

CONCLUSIONS

Combined initial low serum cholesterol levels and obesity appear to indicate a four times greater risk for colon cancer in men as compared with people with average values of both variables. The reasons for these observations are unknown.

Colon cancer overview

The scope of current prevention research support by the National Cancer Institute includes the clinical assessment of dietary modifications and cancer screening trials, epidemiologic studies, development of new chemopreventive therapies, and the use of advanced molecular biologic technologies to probe the genetic determinants of colorectal adenomas. Colorectal cancer frequently has been associated with high-fat low-fiber diets in epidemiologic and experimental studies. A recently initiated Phase III Dietary Intervention Study of Recurrence of Large Bowel Adenomatous Polyps will investigate the potential benefits of a low-fat high-fiber fruit-and-vegetable-enriched eating pattern to decrease the polyp recurrence rate. The Chemoprevention Program currently is supporting four Phase III controlled clinical intervention trials investigating the cancer-inhibiting effects on colorectal cancer of beta-carotene, piroxicam, calcium, and calcium plus fiber in persons with previous adenomas. A proposed early detection trial will screen for colorectal, prostate, lung, and ovarian cancers. A comparison of incidence and mortality trends indicates progress in colorectal cancer detection and therapy.

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.

Obesity in youth and middle age and risk of colorectal cancer in men

To investigate an association between colon cancer and obesity during early adulthood--a potentially important period in the etiology of this disease--the authors assembled, by computer linkage, a population-based historical cohort of 52,539 men born between 1913 and 1927 residing in Hawaii (USA), for whom weight and height had been recorded in 1942-43 and 1972. Linkage of this cohort to the Hawaii Tumor Registry resulted in the identification of 737 incident cases of colorectal cancer for 1972-86. An average of 3.8 cancer-free controls were matched to each case on month and year of birth and ethnicity of the parents. A case-control analysis in each anatomic subsite of the large bowel revealed that both early and middle-age body mass increased the risk of sigmoid cancer in men in a dose-dependent fashion. The odds ratios (OR) for sigmoid cancer for the highest compared with the lowest tertiles of Quetelet index were: 2.1 (95 percent confidence interval [CI] = 1.4-3.2) and 1.7 (CI = 1.1-2.5), at ages 15-29 and in prediagnostic years, respectively. These associations were additive and independent of socioeconomic status. Men who were above the median Quetelet index in 1942 and 1972 had an OR of 2.7 (CI = 1.8-4.0), compared with those who were below the median in both periods. This study provides further evidence for an association of obesity with colon cancer in men and suggests that this association is limited to the sigmoid colon and may be related to both early and late events of colon carcinogenesis.

Capacity of adipose tissue to promote growth and metastasis of a murine mammary carcinoma: effect of estrogen and progesterone

Previously we have shown that a murine mammary carcinoma cell line, designated SPI, grows and metastasizes more efficiently in the mammary gland than in the subcutis. In this report, we examine the tissue specificity of this phenomenon. Our results show that SPI cells grow best in the mesenteric and ovarian fat pads and well in the mammary gland, but very poorly in the subcutis or peritoneal cavity. Massive dissemination of tumors from the ovarian and mesenteric sites occurs to the liver, spleen and diaphragm. In contrast, metastases from the mammary site occur primarily in the lung. Co-transplantation of a threshold number of SPI cells with mammary or ovarian fat fragments into the subcutis results in increased tumor growth, whereas very few tumors form in sham controls receiving no fat fragments. Removal of the ovaries of donor and recipient mice abrogates tumor growth in adipose tissue transplants. Estrogen can stimulate growth of SPI in adipose tissue sites, whereas progesterone inhibits growth. In contrast, in vivo growth of a stable metastatic variant selected from SPI cells was not inhibited by progesterone. SPI cells growing in ovarian and mesenteric fat pads showed increased expression of estrogen receptors and progesterone receptors, as well as detectable levels of epidermal-growth-factor receptors, whereas receptor levels decreased to baseline on tumors in the subcutis. The levels of estrogen-receptor mRNA reflect the corresponding functional expression of receptors; this finding suggests that the regulation of estrogen-receptor expression in this system is, at least in part, at the mRNA level. Our results are consistent with the model that adipose tissue exerts an estrogen-dependent positive regulatory effect on primary SPI tumor growth, and promotes the formation of metastases.

In vitro adsorption of a hydrophobic mutagen to gastrointestinal mucus glycoprotein (mucin) and dietary fibre

The adsorption of mutagens by some dietary fibres has been suggested as one mechanism by which dietary fibres protect against colorectal cancer. It is thought that these dietary fibres carry the mutagen out of the digestive tract, decreasing the effective mutagen concentration to which epithelial cells are exposed. The ability of gastrointestinal mucin to alter the extent to which the hydrophobic mutagen 1,8-dinitropyrene (DNP) adsorbs in vitro onto the insoluble dietary fibre alpha-cellulose, was investigated. It was found that crude and purified human ileal mucins themselves adsorbed DNP and decreased the adsorption of DNP onto alpha-cellulose. Purified mucin which had been treated with trypsin also adsorbed DNP. These studies suggest that in the digestive tract there would be competition for the adsorption of DNP between mucin and insoluble dietary fibres, such as alpha-cellulose. This factor must be considered in predictions about the distribution of hydrophobic, mutagenic carcinogens in the digestive tract and their role in the etiology of colorectal cancer.

Alcohol and colorectal cancer: a case-control study from northern Italy

The role of alcohol consumption in the etiology of colorectal cancer has been investigated in a case-control study conducted from 1985 to 1990 in the northern part of Italy, on 889 cases of colon cancer, 581 cases of rectal cancer, and 2,475 controls admitted to hospital for acute, non-neoplastic, nondigestive disorders. After allowance for age, education, study center, body mass index, and approximate total energy intake, no significant associations between alcohol intake and the risk of cancer of the colon or rectum were found (odds ratios [OR] for greater than or equal to 42 drinks/week cf none = 1.0 (95 percent confidence interval [CI] = 0.8-1.4) and 0.7 (CI = 0.5-1.0) for cancer of the colon and rectum, respectively). A significant increase in the risk of colon cancer with increasing alcohol consumption was, however, observed in females (OR for greater than or equal to 28 drinks/week cf none = 1.8 (CI = 1.1-3.0). While the results of the present case-control study do not suggest that alcohol plays a role in the etiology of colon or rectum cancer overall, they provide a hint for a weak association between alcohol consumption and colon cancer among females which, because of the similarities with breast cancer, should be evaluated in the context of the possible relationship between colon cancer, alcohol intake, and female hormones.

Food consumption and cancer of the colon and rectum in north-eastern Italy

The relation between dietary factors and the risk of colorectal cancer was investigated in a case-control study conducted in Pordenone province, North-eastern Italy, on 123 cases of colon cancer, 125 of rectal cancer and 699 controls admitted to hospital for acute, non-neoplastic or digestive disorders. Consistent positive associations were observed with more frequent consumption of bread (odds ratio, OR = 2.1 for colon and 2.2 for rectum for highest vs. lowest tertile), polenta (OR = 2.1 for colon, 1.9 for rectum), cheese (OR = 1.7 for colon, 1.8 for rectum) and eggs (2.5 for colon, 1.9 for rectum), whereas reduced ORs were observed in subjects reporting more frequent consumption of tomatoes (OR = 0.5 for colon and 0.4 for rectum). High consumption of margarine exerted a significant protection against cancer of the colon whereas high consumption of carrot spinach, whole-grain bread and pasta (favorably) and red meat (unfavorably) affected rectal cancer risk in particular. Thus the present study gives support for a protective effect associated with a fiber-rich or vegetable-rich diet, while it indicates that frequent consumption of refined starchy foods, eggs and fat-rich foods such as cheese and red meat is a risk factor for colo-rectal cancer.

Adsorption of a hydrophobic mutagen to dietary fiber from taro (Colocasia esculenta), an important food plant of the South Pacific

The incidence of colorectal cancer is lower in Polynesian populations of the South Pacific than in European populations. This difference in incidence of the disease may be, at least partly, related to diet. Dietary fiber is believed to protect against colorectal cancer, and one of the ways it may act is by adsorbing mutagens that are carcinogenic. Very little is known about the chemical composition or the ability to adsorb mutagens of these dietary fibers from South Pacific food plants. In contrast to European food plants, which are mostly dicotyledons, South Pacific food plants are mainly monocotyledons. We isolated cell walls (dietary fiber) from the three edible parts of taro (Colocasia esculenta), which is a monocotyledon and a major South Pacific food plant. The ability of these three unlignified cell-wall preparations to adsorb the hydrophobic environmental mutagen 1,8-dinitropyrene was studied. The greatest adsorption occurred with walls from leaf blade, followed by petiole and corm walls, although the differences were not major. The amount of adsorption was intermediate between the low adsorption previously found with unlignified dicotyledon walls (from the flesh of potato tubers and immature cabbage leaves) and the much higher adsorption found with unlignified walls from monocotyledons of the grass and cereal family (Poaceae) (from leaves of seedling Italian ryegrass). These data are consistent with the monosaccharide compositions of the taro wall preparations, which were more similar to those of unlignified walls of dicotyledons than to unlignified walls of the Poaceae. These findings are consistent with the hypothesis that the composition of the dietary fiber determines its adsorptive properties and that there may be important differences between the major dietary fibers of South Pacific and European food plants.

Additional ecological evidence: lipids and breast cancer mortality among women aged 55 and over in China

That dietary fat increases breast cancer risk has been strongly supported by international data collected among developed countries during the past few decades. Population aggregates with elevated lipid intake have tended to report elevated breast cancer incidence and mortality. This study is an ecological analysis of the association of various indicators of lipid intake with breast cancer mortality in 65 county-wide population aggregates in the People's Republic of China. Although the result is consistent with a positive association between lipid intake and breast cancer risk, the observed association is weaker than the association previously observed. This finding provides only modest support for the possibility of a diet-breast cancer link.

Occupational exposures and cancer of the colon and rectum

The associations between occupational risks and colorectal cancer were examined in a Swedish population-based, case-referent study. The study was performed in Stockholm in 1986-1988 and included 569 cases and 512 referents. Relative risks (RRs) with 95% confidence intervals were calculated for different occupations/chemicals. Elevated risks of colon cancer were found among male petrol station/automobile repair workers (RR = 2.3, 0.8-6.6) and among males exposed to asbestos (RR = 1.8, 0.9-3.6), while elevated risks of rectal cancer were found among males exposed to soot (RR = 2.2, 1.1-4.3), asbestos (RR = 2.2, 1.0-4.7), cutting fluids/oils (RR = 2.1, 1.1-4.0), and combustion gases from coal/coke/wood (RR = 1.9, 1.0-3.7). However, due to a high correlation between the above-mentioned variables and the few exposed subjects, it is difficult to separate their effects properly. These analyses were adjusted for age. Further adjustments for diet, body mass, and physical activity had little or no influence on the results.

No effect of adult dietary fat on tumors induced prenatally by diethylstilbestrol

Strain CD-1 female mice exposed prenatally to diethylstilbestrol (DES) or vehicle were placed on semipurified diets containing 2.6%, 10%, 20%, or 29% fat by weight at four weeks of age. These mice were used as a breeding colony for a few weeks and then maintained to terminal illness on the semipurified diets. Females exposed prenatally to DES developed mammary tumors, pituitary tumors, and glandular tumors of the reproductive tract. There was no significant difference in tumor frequency between low- and high-fat dietary groups. Fewer tumors appeared in the vehicle-exposed mice, as expected, and their frequency did not differ between the dietary groups. Pregnancy reduced tumor frequency in DES-exposed mice, but the incidence of pregnancy was not significantly different between low- and high-fat dietary groups. In the adult the failure of a high-fat diet to increase the frequency of reproductive system tumors induced prenatally is in marked contrast to the effectiveness of high-fat diets in promoting mammary tumors induced by carcinogens given to rats postnatally. This difference is critical in the interpretation of epidemiological studies. The relationship of dietary fat to reproductive system cancer in human populations was reviewed in comparison with these two animal models. The epidemiological literature was found to be more consistent with the animal model, showing high sensitivity to dietary fat prenatally but no significant sensitivity at the adult stage of life.

Mutations in p53 as potential molecular markers for human breast cancer

Based on the high incidence of loss of heterozygosity for loci on chromosome 17p in the vicinity of the p53 locus in human breast tumors, we investigated the frequency and effects of mutations in the p53 tumor suppressor gene in mammary neoplasia. We examined the p53 gene in 20 breast cancer cell lines and 59 primary breast tumors. Northern blot analysis, immunoprecipitation, and nucleotide sequencing analysis revealed aberrant mRNA expression, over-expression of protein, and point mutations in the p53 gene in 50% of the cell lines tested. A multiplex PCR assay was developed to search for deletions in the p53 genomic locus. Multiplex PCR of genomic DNA showed that up to 36% of primary tumors contained aberrations in the p53 locus. Mutations in exons 5-9 of the p53 gene were found in 10 out of 59 (17%) of the primary tumors studies by single-stranded conformation polymorphism analysis. We conclude that, compared to amplification of HER2/NEU, MYC, or INT2 oncogene loci, p53 gene mutations and deletions are the most frequently observed genetic change in breast cancer related to a single gene. Correlated to disease status, p53 gene mutations could prove to be a valuable marker for diagnosis and/or prognosis of breast neoplasia.

Proliferative potential and expression of cell type specific functions in primary mouse colonic epithelial cells

Primary cultures of mouse colonic epithelial cells have been obtained that are typically epithelial by morphology and moreover express keratins and endogenous beta-galactosidase; this latter activity was also demonstrated in the epithelial lining of the mouse colonic mucosa. The proliferative response of the primary colonic epithelial cells to epidermal growth factor, insulin, and the bile acid, deoxycholic acid, has been studied. Using primary cultures maintained at suboptimal growth conditions, which yielded 96 to 100% quiescent cells, epidermal growth factor, insulin, and the bile acid, deoxycholic acid, at concentrations at which it normally occurs in the aqueous phase of human feces, stimulated proliferation as measured by autoradiography. Exposure of the cells to combinations of these factors resulted in additive increases in growth. In conclusion, cells from the normal mouse colon can now be cultured while retaining at least two normal marker functions and moreover respond to some known mitogens and the potential tumor promoter deoxycholic acid. The cells can also be subcultivated while maintaining their epithelial morphology and marker functions for at least 3 passages.

Colon cancer incidence: recent trends in the United States

Between 1976-78 and 1985-87, the age-adjusted incidence rates of invasive colon cancer in the United States rose by 15 percent, 3 percent, 21 percent, and 16 percent among White males, White females, Black males, and Black females, respectively. The increases in incidence occurred in all age groups over age 54 and affected each of the major subsites of the colon nearly equally. The larger rates of increase have resulted in higher incidence among Blacks than Whites by the mid-1980s and an increasingly greater excess of this cancer in males. Trends toward earlier diagnosis of invasive colon cancer were found, with increasing rates for localized and regional diseases coupled with stable or decreasing distant-stage disease-rates. The incidence of in situ colon cancer also rose substantially. The findings suggest that changes in diagnostic trends and risk-factor prevalence may be contributing to these patterns, and that the era when colon cancer predominated among White females is clearly over.