Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer Int. J. Cancer: Supplement 10, 13–17 (1997) r 1997 Wiley-Liss, Inc. DIET, NUTRITION AND PROSTATE CANCER Graham GILES* and Paul IRELAND Cancer Epidemiology Centre, Anti-Cancer Council of Victoria, Carlton South, Victoria, Australia Although much has been written, little is known about the causes of prostate cancer. Variations between populations in the incidence of invasive cancers, together with changes in the incidence of invasive cancers in migrants, have pointed to environmental (lifestyle) factors that might be amenable to intervention. Conversely, there is a lack of international variation in the prevalence of microscopic tumours, so the essential question is: what causes only some of the common microscopic tumours to become aggressive? Dietary factors hold the most promise in this regard and have been the subject of recent reviews. The strongest and most consistent effects are positive associations with animal products such as red meats, eggs and dairy foods, and possibly by implication, fat. Evidence of a protective effect of fruit and vegetables is weak and inconsistent, as is the relationship with vitamin A and carotenoids, such as b-carotene. There are some interesting leads. Lycopene, the carotenoid found in tomatoes, has been reported to be protective; a-tocopherol supplementation has shown a protective effect in one intervention study; and vitamin D has been shown to be protective in a prospective study. Interest is also growing in phytoestrogens and the extent to which dietary manipulation with these and other phytochemicals might influence prostate cancer by modifying male sex hormone levels or actions. There is limited evidence of associations with obesity. It is not known whether these are related to a particular dietary pattern or to possible physiological effects on the male’s hormonal milieu. Associations with lean body mass are likely to be related to the action of androgens during growth and development. Dietary and nutritional effects on prostate cancer do not appear to be strong, but they may be subtle and attenuated by measurement error. To explore these aspects further will require large prospective studies that include improved (repeated) dietary measurements and also blood sampling, so that genetic polymorphisms can be adequately investigated. Such studies are underway. Int. J. Cancer Suppl. 10:13–17, 1997. r 1997 Wiley-Liss, Inc. Prostate cancer is the leading cancer in males of many countries. Little is known of its causes and nothing can be done to prevent it. Prostate cancer epidemiology, particularly in regard to diet, presents interesting challenges. In its microscopic histological form, prostate cancer attains a very high prevalence with increasing age, but in few men does the small focus of cancer cells become invasive. Many prostate ‘‘cancers’’ recorded by cancer registries are diagnosed incidentally during treatment for benign enlargement or by screening using blood PSA levels. The observation that the microscopic form (often termed ‘‘latent’’) is of equivalent prevalence in both high- and low-risk populations for the invasive form of the disease supports the probability of a lifestyle exposure being implicated in progression (Yatani et al., 1982). Further evidence of environmental causes comes from the observation of marked international variation and increased incidence in migrant populations from low- to high-risk communities. In Italian migrants to Australia, for example, the risk of mortality from prostate cancer increases with duration of residence, but even after more than 30 years remains less than the risk to the Australian-born (Geddes and Buiatti, 1993). Of the environmental and lifestyle factors that have been studied, diet offers the most interesting leads, and has been the subject of much epidemiological investigation. Ecological studies (Blair and Fraumeni, 1978; Howell, 1974; Armstrong and Doll, 1975; Rose et al., 1986) have shown striking correlations between international prostate cancer mortality and per capita consumption of fat. Measurement of diet in epidemiological studies at the level of the individual has been extremely varied, and even the better ‘‘vali- dated’’ methods perform badly. Because diet is poorly measured, very large studies are needed to detect statistically meaningful effects. The earlier generation of studies that examined diet and prostate cancer were invariably small, used rudimentary dietary instruments and further, were usually retrospective, relying on recall. This has resulted in a confused picture of weak and inconsistent findings. METHODS Two recent reviews (Key, 1995; Kolonel, 1996) were used as the basis of a bibliographical search. Further studies were identified from computer searches of standard bibliographical databases, but were limited to reports in English. Studies were restricted to those that measured dietary intakes or anthropometry; those based solely on serum micronutrient concentrations were excluded from the overview. Estimates of risk associated with dietary and anthropometric variables were extracted, with confidence intervals when available. Care was taken to include null findings that were reported without formal statistical estimates. Because the odds ratios (ORs) and relative risks (RRs) were calculated for a variety of different exposures, used a range of cut points and were often adjusted for one or more factors, the unadjusted estimates were usually chosen. Due to the heterogeneity of study design, exposure measurement and analysis, no formal summary estimate of effect was attempted. Measures of dietary exposure varied considerably in scope and quality between studies. For this reason, the measures were grouped into broad categories: meat, fat, alcohol, vegetables and fruit. The inclusion of a study was sometimes based on a single vegetable or meat (beef). The nutritional covariate most commonly reported was obesity described by body mass index (BMI). RESULTS The majority of reports available for overview originated from 36 retrospective and 16 prospective studies. The studies are not referenced here, but are included in the References. Only 2 of the 36 retrospective studies contained over 400 cases. Of the 16 prospective studies, only 3 contained over 400 cases. Further, the methods used to collect dietary information were often rudimentary and/or described in vague terms. Few studies combined adequate sample size and good dietary assessments. Recently, studies have begun to address disease status of cases and controls. Many have included clinically insignificant cancers in their case series and, in recent times, will have included screen-detected cancers of unknown clinical significance. These factors ensure that a significant proportion of the findings discussed below will be spurious. The overview is summarised in Table I. This lists the number of retrospective and prospective studies separately and gives the number of positive (OR or RR $ 1.1) and negative (OR or RR # 0.9) associations reported. The number of statistically significant ORs or RRs is given in parentheses, and the range of significant ORs and RRs is given to identify the strength of effect. Because of subgroup analysis, the sum of associations is sometimes greater than the number of studies. Fat Dietary fat appears to be the strongest environmental risk factor for prostate cancer yet identified. Even so, the evidence is *Correspondence to: Cancer Epidemiology Centre, 1 Rathdowne Street, Carlton South, VIC 3053, Australia. Fax: 61-3-9279-1210. e-mail: firstname.lastname@example.org GILES AND IRELAND 14 TABLE I – SUMMARY OF REPORTED ASSOCIATIONS (ODDS RATIOS/RELATIVE RISKS) WITH DIETARY INTAKES AND OBESITY Retrospective studies Studies Total fat Animal fat Meat Dairy products Eggs Fruit Vegetables Vitamin A/retinol b-carotene Alcohol Obesity 14 10 10 7 7 2 7 8 6 9 17 Prospective studies Associations Positive1 Negative2 Range3 12 (3) 12 (3) 11 (3) 4 (4) 6 (1) 2 (0) 4 (0) 4 (0) 3 (0) 5 (1) 7 (2) 6 (0) 6 (0) 0 (0) 0 (0) 1 (0) 1 (1) 12 (2) 2 (1) 4 (0) 5 (0) 0 (0) 1.4–1.7 1.7–2.8 1.7–2.6 1.6–3.1 2.1 0.4 0.3–0.5 0.4 1.9 4.0–4.4 Studies 2 3 10 7 5 8 6 1 3 6 10 Associations Positive1 Negative2 1 (0) 4 (1) 7 (3) 3 (2) 3 (0) 2 (0) 5 (0) 1 (1) 1 (1) 2 (0) 8 (3) 1 (0) 0 (0) 3 (0) 2 (0) 2 (0) 2 (0) 4 (2) 1 (1) 1 (1) 1 (0) 1 (0) Range3 1.6 1.6–2.6 1.4–2.4 0.6–0.7 0.4–2.8 0.2–1.9 1.3–2.4 ORs/RRs $ 1.1 (numbers in parentheses are statistically ORs/RRs #0.9 (numbers in parentheses are statistically significant).–3Range, lowest to highest statistically significant ORs/RRs. 1Positive, inconclusive. Fat intake is often implied from certain food sources that are high in fat, and it is often difficult to discern whether fat per se or some other property of these foods is the important exposure. A total of 16 studies reported total fat intake. These yielded 13 OR/RR estimates $1.1 and 6 estimates #0.9. Only 3 were statistically significant, ranging from 1.4–1.7. Of the 13 studies that measured fat from animal sources, 16 had OR/RR estimates $1.1 and 6 #0.9. Only 4 estimates were statistically significant, ranging from 1.6–2.8. Some studies (Kolonel et al., 1983, 1988; West et al., 1991) reported a different effect of fat with age, with low risks under age 70 and increased risks over age 70. Animal foods Meat is a fairly heterogeneous entity when measured in epidemiological studies. We concentrated on beef in this overview. Its relationship to cancer risk could include the following: an effect of meat alone, an effect of the fat contained in meat, or an effect of carcinogens created in cooking or some other correlate of meat, e.g., low vegetable intake. There is no convincing evidence to support any of these hypotheses. Of the 20 studies included, 18 gave OR/RR estimates $1.1 and 3 #0.9. Of these, 6 OR/RR estimates were statistically significant, ranging from 1.6–2.6. The highest OR of 2.6 was reported by Bravo et al. (1991). The highest RR was 2.64 (Giovannucci et al., 1993); after multivariate adjustment this was reduced to 1.9. Fourteen studies measured intake of dairy products. These produced 7 OR/RR estimates $1.1 and 2 #0.9. There were 4 ORs significantly $1.1 (range, 1.6–3.1) and 2 RRs $1.1 (range, 1.4–2.4). Associations with eggs were reported in 12 studies. Of these, 9 estimates were $1.1 and 3 were #0.9. Only one OR was statistically significant at 2.1 (Walker et al., 1992). Plant foods Ten studies, 2 retrospective and 8 prospective, reported on fruit consumption. Of these, 3 were negative and 4 were positively associated with risk. Only one retrospective study (Negri et al., 1991) gave a statistically significant OR of 0.4. Vegetables were measured in various ways in different studies. Altogether, 13 studies reported on at least one vegetable or vegetable group. These yielded 9 positive associations and 16 negative associations that included 4 significant OR/RRs. Carrots were found by Walker et al. (1992) to give an OR of 0.5; green vegetables were found by Negri et al. (1991) to give an OR of 0.3; and tomato consumption was found by Mills et al. (1989) and Giovannucci et al. (1996) to give RRs of 0.6 and 0.65, respectively. Dietary b-carotene was estimated in 9 studies, 4 of which gave estimates $1.1 and 5 #0.9. Only one study (Hsing et al., 1990) significant).–2Negative, yielded significant results. In this prospective study, men ,75 years old had increased risk associated with b-carotene, RR 1.9 (1.0– 3.7), and men older than 75 had reduced risk, RR 0.2 (0.1–0.6). Alcohol Alcohol consumption appears to have little association with prostate cancer risk. Of 8 earlier retrospective studies that included alcohol, none gave an OR that was significantly different from 1.0. A similar pattern was observed in the 6 prospective studies. However, Hayes et al. (1996) published an OR of 1.9 (1.3–2.7) from a large case-control study. This study was unusual in that elevated risk was observed at very high levels of alcohol consumption. Obesity Obesity was reported in most studies as BMI. Of the 17 reports from retrospective studies, only the study of Talamini et al. (1986) yielded a statistically elevated OR of 4.0 that increased a little after adjustment to 4.4. For comparison, there were 10 prospective studies that reported associations with BMI. Of these, 8 OR/RR estimates were $1.1 and one was #0.9. Three of the elevated RRs were statistically significant, the 2 largest being 2.4 and 2.5 (adjusted) from Snowdon et al. (1984). For a measure that is fairly accurate, the BMI results are fairly inconsistent. This may be due in part to the failure of BMI to distinguish between lean and fat mass. Lean mass is related to androgen supply and may be positively associated with prostate cancer risk (Severson et al., 1988). On the other hand, fat mass is a source of androstenedione that can be aromatised to form oestrogens that might be protective against prostate cancer. Studies that have examined height (Kolonel et al., 1988; Whittemore et al., 1995; La Vecchia et al., 1990; Severson et al., 1988) have found no associations. There has been one report of an association with high birth weight (Tibblin et al., 1995). Other foods and nutrients Dietary vitamin A was estimated in various ways in 9 studies, of which 3 gave negative and 6 gave positive associations. Only 2 studies produced statistically significant estimates. The case control study of Negri et al. (1991) gave an OR of 0.4. Hsing et al. (1990) in a prospective study showed a similar effect to that for b-carotene. Men below age 75 had an increased RR of 2.8, and men older than 75 had reduced RR of 0.4. A protective effect of a-tocopherol has been observed in one intervention trial (ATBC Cancer Prevention Group, 1994), suggesting that this micronutrient might be worth following up in other studies. However, concentrations of vitamin D [1,25-(OH)2 D in serum] shown to be protective DIET, NUTRITION AND PROSTATE CANCER in a prospective study (Corder et al., 1993) have not been confirmed in another (Braun et al., 1995). In summary, much of the evidence of associations between diet and prostate cancer is weak and inconsistent. This is related to the small size and quality of many of the retrospective studies, and poor dietary measurement in some of the prospective studies. The data from the better prospective studies suggest that animal foods/fat and carotenoids are significantly associated with prostate cancer risk and, further, possibly interact with each other and with age. These aspects merit further follow-up in prospective cohort studies that use validated dietary instruments, and that are of a size to provide sufficient statistical power to examine interactions. Ideally, research should be performed in populations that demonstrate a wide range in levels of intake. FUTURE RESEARCH It is important that the trend towards specifying the nature of cases of prostate cancer is continued and improved. We are interested in discovering the causes of clinically significant disease rather than the cause of common microscopic foci of cancer. To this end, it might be useful to collect blood and tissue samples against the day when molecular and genetic markers will be available to explore this avenue further. We do not know, for example, how dietary intakes might interact with genetic factors that control the androgen receptor or 5 a-reductase. There is probably little to be gained from conducting more small case-control studies of diet and prostate cancer that measure diet poorly. When the current generation of case-control studies is completed, it will be time to plan new research that has newly refined objectives and methods. There are now several large prospective cohorts of men that will yield considerable numbers of prostate cancer over the next 5–10 years. Provided that these studies have measured diet in a reasonable way, they may offer more evidence of effects. Better dietary assessment will continue to be a long-term goal. In cohort settings, advances in dietary assessment methods can be built fairly easily into repeated measures made at periodic follow-ups. It is also important that such studies collect anthropometric measures of lean and fat mass to be able to qualify our understanding of the risks associated with obesity. There remains an important role for ecological studies in contemporary nutritional epidemiology. While it is true that ecological studies are most appropriate for generating hypotheses which then can be tested in case-control and cohort studies, the extent to which ecological studies are being used for this purpose in current prostate cancer epidemiology is not evident. Foods that are only now being recognised as potentially protective for prostate cancer could have been identified as useful leads from comparison of high- and low-risk populations several decades ago. Tomato use 15 in southern Europe and soya bean products in Japan are just 2 examples. We need to broaden our horizons when thinking about diet. While nutrients may be important determinants of cancer risk, there are many other possibilities in terms of foods, non-nutritive components of foods and food preparation practices. Australian contributions Two studies, a large case-control study and a prospective cohort study, are underway in Australia that are designed to overcome some of the shortcomings of previous efforts (Giles, 1990). The case-control study is being conducted in Melbourne, Sydney and Perth and will enroll over 2,000 cases and an equal number of age-frequency-matched population-based controls. Cases are restricted to adenocarcinomas of moderate to poor differentiation, and we are also collecting staging data and information on the mode of detection. We are collecting blood from every case and from a sample of controls. The dietary instrument is a 121-item FFQ that was developed for the cohort study, and is used in combination with graded pictures of food portions obtained from pilot studies of weighed food records from over 900 Melburnians. This instrument has been validated in a sample of the cohort study and performs comparatively well (Ireland, 1996). For example, we have correlated its estimate of protein intake against urinary nitrogen output and have obtained a Pearson’s R of 0.25 for males and 0.37 for females. This compares favourably with the correlations for females found when validating the Oxford and Cambridge dietary instruments used in the EPIC cohort studies (0.24 and 0.15, respectively) (Bingham et al., 1995). In our Australian cohort of 17,000 men, we used the dietary instruments described above and have also measured anthropometric indices, including body impedance, in order to explore the question of obesity associations being related to lean or fat mass. The cohort was aged 40–69 at recruitment; they are now 5 years older. Men of southern European birth (Italy, Greece and Malta), who were deliberately oversampled to add dietary range to the study population, represent 28% of the sample. Extensive questionnaire data were collected on lifestyle exposures, tobacco, alcohol, physical activity and other factors. A blood sample was collected from each individual, and aliquots were stored in liquid nitrogen and as dried blots on Guthrie cards. The study centre also runs the population-based cancer registry and routinely links with the death register to identify endpoints. Together, these studies address the issues identified with disease characterisation, dietary assessment, anthropometry, tissue collection, population sampling and statistical power. In regard to the ability to further the dietary investigation of some of the newer micronutrients of interest, they are currently limited by the availability of food tables for many phytochemicals, including carotenoids and phytoestrogens. REFERENCES ALPHA-TOCOPHEROL, BETA-CAROTENE CANCER PREVENTION STUDY GROUP, The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. N. Engl. J. Med., 330, 1029–1035 (1994). ARMSTRONG, B. and DOLL, R., Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices. Int. J. Cancer, 15, 617–631 (1975). BINGHAM, S.A., CASSIDY, A., COLE, T.J., WELCH, A., RUNSWICK, S.A., BLACK, A.E., THURNHAM, D., BATES, C., KHAW, K.T., KEY, T.J. and DAY, N.E., Validation of weighed records and other methods of dietary assessment using the 24 hour urine nitrogen technique and other biological markers. Brit. J. Nutr., 73, 531–550 (1995). BLAIR, A. and FRAUMENI, J.F., JR., Geographic patterns of prostate cancer in the United States. J. nat. Cancer Inst., 61, 1379–1384 (1978). BRAUN, M.M., HELZLSOUER, K.J., HOLLIS, B.W. and COMSTOCK, G.W., Prostate cancer and prediagnostic levels of serum vitamin D metabolites. Cancer Causes Control, 2, 235–239 (1995). BRAVO, M.P., CASTELLANOS, E. and DEL REY CALERO, J., Dietary factors and prostatic cancer. Urol. Int., 46, 163–166 (1991). CORDER, E.H., GUESS, H.A., HULKA, B.S., FRIEDMAN, G.D., SADLER, M., VOLLMER, R.T., LOBAUGH, B. and DREZNER, M.K., Vitamin D and prostate cancer: a prediagnostic study with stored sera. Cancer Epidemiol. Biomarkers Prev., 2, 467–472 (1993). DEMARK-WAHNEFRIED, W., PAULSON, D.F., ROBERTSON, C.N., ANDERSON, E.E., CONAWAY, M.R. and RIMER, B.K., Body dimension differences in men with or without prostate cancer. J. nat. Cancer Inst., 84, 1363–1364 (1992). FINCHAM, S.M., HILL, G.B., HANSON, J. and WIJAYASINGHE, C., Epidemiology of prostatic cancer: a case-control study. Prostate, 17, 189–206 (1990). GANN, P.H., HENNEKENS, C., SACKS, F.M., GRODSTEIN, F., GIOVANNUCCI, E.L. and STAMPFER, M.J., Prospective study of plasma fatty acids and risk of prostate cancer. J. nat. Cancer Inst., 86, 281–286 (1994). 16 GILES AND IRELAND GARFINKEL, L., Overweight and mortality. Cancer, 58, 1826–1829 (1986). GEDDES, M. and BUIATTI, E., Review of studies of cancer in Italian migrants. In: M. Geddes, D.M. Parkin, M. Khlat, D. Balzi and E. Buiatti (eds.), Cancer in Italian Migrant populations, IARC scientific Publication 123, pp. 24–30, Lyon (1993). GHADIRIAN, P., LACROIX, A., MAISONNEUVE, P., PERRET, C., DROVIN, G., PERRAULT, J.P., BÉLAND, G., ROHAN, T.E. and HOWE, G.R., Nutritional factors and prostate cancer: a case-control study of French Canadians in Montreal, Canada. Cancer Causes Control, 7, 428–436 (1996). GILES, G.G., The Melbourne study of diet and cancer. Proc. Nutr. Soc. Aust., 15, 61–68 (1990). GIOVANNUCCI, E., RIMM, E.B., COLDITZ, G.A., STAMPFER, M.J., ASCHERIO, A., CHUTE, C.C. and WILLETT, W.C., A prospective study of dietary fat and risk of prostate cancer. J. nat. Cancer Inst., 85, 1571–1579 (1993). GIOVANNUCCI, E., ASCHERIO, A., RIMM, E.B., STAMPFER, M.J., COLDITZ, G.A. and WILLETT, W., Intake of carotenoids and retinol in relation to risk of prostate cancer. J. nat. Cancer Inst., 87, 1767–1776 (1996). GRAHAM, S., HAUGHEY, B., MARSHALL, J., PRIORE, R., BYERS, T., RZAPKA, T., METTLIN, C. and PONTES, J.E., Diet in the epidemiology of carcinoma of the prostate gland. J. nat. Cancer Inst., 70, 687–692 (1983). GREENWALD, P., DAMON, A., KIRMSS, V. and POLAN, A.K., Physical and demographic features of men before developing cancer of the prostate. J. nat. Cancer Inst., 53, 341–346 (1974). GRONBERG, H., DAMBER, L. and DAMBER, J.E., Total food consumption and body mass index in relation to prostate cancer risk: a case-control study in Sweden with prospectively collected exposure data. J. Urol., 155, 969–974 (1996). HAYES, R.B., DE JONG, F.H., RAATGEVER, J., BOGDANOVICZ, J., SCHROEDER, F.H., VAN DER MAAS, P., OISHI, K. and YOSHIDA, O., Physical characteristics and factors related to sexual development and behaviour and the risk for prostatic cancer. Europ. J. Cancer Prev., 1, 239–245 (1992). HAYES, R.B., BROWN, L.M., SCHOENBERG, J.B., GREENBERG, R.S., SILVERMAN, D.T., SCHWARTZ, A.G., SWANSON, G.M., BENIDOU, J., LIFF, J.M., HOOVER, R.N. and POTTERN, L.M., Alcohol use and prostate cancer risk in US blacks and whites. Amer. J. Epidemiol., 143, 692–697 (1996). HESHMAT, M.Y., KAUL, L., KOVI, J., JACKSON, M.A., JACKSON, A.G., JONES, G.W., EDSON, M., ENTERLINE, J.P., WORRELL, R.G. and PERRY, S.L., Nutrition and prostate cancer: a case-control study. Prostate, 6, 7–17 (1985). HIATT, R.A., ARMSTRONG, M.A., KLATSKY, A.L. and SIDNEY, S., Alcohol consumption, smoking, and other risk factors and prostate cancer in a large health plan cohort in California (United States). Cancer Causes Control, 5, 66–72 (1994). HIETANEN, E., BARTSCH, H., BERZIAT, J., CAMUS, A., MCCLINTON, S., EREMIN, O., DAVIDSON, S.L. and BOYLE, P., Diet and oxidative stress in breast, colon and prostate cancer patients: a case-control study. Europ. J. clin. Nutr., 48, 575–586 (1994). HIRAYAMA, T., Epidemiology of prostate cancer with special reference to the role of diet. J. nat. Cancer Inst., 53, 149–155 (1979). HONDA, G.D., BERNSTEIN, L., ROSS, R.K., GREENLAND, S., GERKINS, V. and HENDERSON, B.E., Vasectomy, cigarette smoking, and age at first sexual intercourse as risk factors for prostate cancer in middle-aged men. Brit. J. Cancer, 57, 326–331 (1988). HOWELL, M.A., Factor analysis of international cancer mortality data and per capita food consumption. Brit. J. Cancer, 29, 328–336 (1974). HSING, A.W., MCLAUGHLIN, J.K., SCHUMAN, L.M., BJELKE, E., GRIDLEY, G., WACHOLDER, S., CHIEN, H.T. and BLOT, W.J., Diet, tobacco use, and fatal prostate cancer: results from the Lutheran Brotherhood cohort study. Cancer Res., 50, 6836–6340 (1990). IRELAND, P.D., Measuring dietary intake in a Melbourne cohort of men and women born in Australia, Italy and Greece. Doctoral thesis, Department of Medicine, Monash University, Clayton, Victoria, Australia (1996). KAUL, L., HESHMEAT, M.Y., KOVI, J., JACKSON, M.A., JACKSON, A.G., JONES, G.W., EDSON, M., ENTERLINE, J.P., WORRELL, P.G. and PERRY, S.L., The role of diet in prostate cancer. Nutr. Cancer, 9, 123–128 (1987). KEY, T., Risk factors for prostate cancer. Cancer Surv., 23, 63–77 (1995). KOLONEL, L.N., Nutrition and prostate cancer. Cancer Causes Control, 7, 83–94 (1996). KOLONEL, L.N., HANKIN, J.H., LEE, J., CHU, S.Y., NOMURA, A.M.Y. and HINDS, M.W., Nutrient intakes in relation to cancer incidence in Hawaii. Brit. J. Cancer, 44, 332–339 (1983). KOLONEL, L.N., HANKIN, J.H. and YOSHIZAWA, C.N., Vitamin A and prostate cancer in elderly men: enhancement of risk. Cancer Res., 47, 2982–2985 (1987). KOLONEL, L.N., YOSHIZAWA, C.N. and HANKIN, J.H., Diet and prostatic cancer: a case-control study in Hawaii. Amer. J. Epidemiol., 127, 999–1012 (1988). LA VECCHIA, C., NEGRI, E., PARAZZINI, F., BOYLE, P., D’AVANZO, P., LEVI, F., GENTILE, A. and FRANCESCHI, S., Height and cancer risk in a network of case-control studies from northern Italy. Int. J. Cancer, 45, 275–279 (1990). LA VECCHIA, C., NEGRI, E., D’AVANZO, B., FRANCESCHI, S. and BOYLE, P., Dairy products and the risk of prostatic cancer. Oncology, 48, 406–410 (1991). LE MARCHAND, L., HANKIN, J.H., KOLONEL, L.N. and WILKENS, L.R., Vegetables and fruit consumption in relation to prostate cancer risk in Hawaii: a reevaluation of the effect of dietary beta-carotene. Amer. J. Epidemiol., 133, 215–219 (1991). LEMARCHAND, L., KOLONEL, L.N., WILKENS, L.R., MYERS, B.C. and HIROHATA, T., Animal fat consumption and prostate cancer: a prospective study in Hawaii. Epidemiology, 5, 276–282 (1994). MEIKEL, A., STANISH, W. and WEST, D., Risk factors for prostate cancer. Abstracts of the American Public Health Association meeting, Montreal, Canada, November 14–18 (1982). METTLIN, C., SELENSKAS, S., NATARAJAN, N. and HUBEN, R., Beta-carotene and animal fats and their relationship to prostate cancer risk. A case-control study. Cancer, 64, 605–612 (1989). MIDDLETON, B., BYERS, T., MARSHALL, J. and GRAHAM, S., Dietary vitamin A and cancer: a multisite case-control study. Nutr. Cancer, 8, 107–116 (1986). MILLS, P.K., BEESON, L., PHILLIPS, R.L. and FRASER, G.E., Cohort study of diet, lifestyle, and prostate cancer in Adventist men. Cancer, 64, 598–604 (1989). MISHINA, T., WATANABE, H., ARAKI, H. and NAKAO, M., Epidemiological study of prostatic cancer by matched-pair analysis. Prostate, 6, 423–436 (1985). MØLLER, H., MELLEMGAARD, A., LINDVIG, K. and OLSEN, J.H., Obesity and cancer risk: a Danish record-linkage study. Europ. J. Cancer, 30A, 344–350 (1994). NEGRI, E., LA VECCHIA, C., FRANCESCHI, S., D’AVANZO, B. and PARAZZINI, F., Vegetable and fruit consumption and cancer risk. Int. J. Cancer, 48, 350–354 (1991). OHNO, Y., YOSHIDA, O., OISHI, K., OKADA, K., YAMABE, H. and SCHROEDER, F.H., Dietary beta carotene and cancer of the prostate: a case-control study in Kyoto, Japan. Cancer Res., 48, 1331–1336 (1988). PAGANINI-HILL, A., CHAO, A., ROSS, R.K. and HENDERSON, B.E., Vitamin A, beta-carotene and the risk of cancer: a prospective study. J. nat. Cancer Inst., 79, 443–448 (1987). ROHAN, T.E., HOWE, G.R., BURCH, J.D. and JAM, M., Dietary factors and risk of prostate cancer: a case-control study in Ontario, Canada. Cancer Causes Control, 6, 145–154 (1995). ROSE, D.P., BOYAR, A.P. and WYNDER, E.L., International comparisons of mortality rates for cancer of the breast, ovary, prostate, and colon, and per capita food consumption. Cancer, 58, 2363–2371 (1986). ROSS, R.K., SHIMIZU, H., PAGANINI-HILL, A., HONDA, G. and HENDERSON, B.E., Case-control studies of prostate cancer in blacks and whites in southern California. J. nat. Cancer Inst., 78, 869–874 (1987). ROTKIN, I.D., Studies in the epidemiology of prostatic cancer: expanded sampling. Cancer Treat. Rep., 61, 173–180 (1977). SCHUMAN, L.M., MANDEL, J., BLACKARD, C., BAUER, H., SCARLETT, J. and MCHUGH, R., Epidemiologic study of prostatic cancer: preliminary report. Cancer Treat. Rep., 61, 181–186 (1977). SCHUMAN, L.M., MANDEL, J.S., RADKE, A., SEAL, U. and HALBERG, F., Some selected features of the epidemiology of prostatic cancer: Minneapolis-St. Paul, Minnesota case-control study, 1976–1979. In: K. Magnus (ed.), Trends in cancer incidence: causes and implications, pp. 345–354, Hemisphere, Washington, DC (1982). SEVERSON, R.K., GROVE, J.S., NOMURA, A.M.Y. and STEMMERMANN, G.N., Body mass and prostatic cancer: a prospective study. Brit. med. J., 297, 713–715 (1988). SEVERSON, R.K., NOMURA, A.M.Y., GROVE, J.S. and STEMMERMAN, G.N., A prospective study of demographics, diet, and prostate cancer among men of Japanese ancestry in Hawaii. Cancer Res., 49, 1857–1860 (1989). SHIBATA, A., PAGANINI-HILL, A., ROSS, R.K. and HENDERSON, B.E., Intake of vegetables, fruits, beta-carotene, vitamin C and vitamin supplements and cancer incidence among the elderly: a prospective study. Brit. J. Cancer, 66, 673–679 (1992). SLATTERY, M.L. and WEST, D.W., Smoking, alcohol, coffee, tea, caffeine, DIET, NUTRITION AND PROSTATE CANCER and theobromine: risk of prostate cancer in Utah (United States). Cancer Causes Control, 4, 559–563 (1993). SNOWDON, D.A., PHILLIPS, R.L. and CHOI, W., Diet, obesity, and risk of fatal prostate cancer. Amer. J. Epidemiol., 120, 244–250 (1984). TALAMINI, R., LA VECCHIA, C., DECARLI, A., NEGRI, E. and FRANCESCHI, S., Nutrition, social factors and prostatic cancer in a Northern Italian population. Brit. J. Cancer, 53, 817–821 (1986). TALAMINI, R., FRANCESCHI, S., LA VECCHIA, C., SERRAINO, D., BARRA, S. and NEGRI, E., Diet and prostatic cancer: a case-control study in Northern Italy. Nutr. Cancer, 18, 277–286 (1992). TAVANI, A., NEGRI, E., FRANCESCHI, S., TALAMINI, R. and LA VECCHIA, C., Alcohol consumption and risk of prostate cancer. Nutr. Cancer, 21, 25–31 (1994). THOMPSON, M.M., GARLAND, C., BARRETT-CONNOR, E., KHAW, K., FRIEDLANDER, N.J. and WINGARD, D.L., Heart disease risk factors, diabetes and prostatic cancer in an adult community. Amer. J. Epidemiol., 129, 511–517 (1989). TIBBLIN, G., ERIKSSON, M., CNATTINGIUS, S. and EKBOM, A., High birthweight as a predictor of prostate cancer risk. Epidemiology, 6, 423–424 (1995). WALKER, A.R.P., WALKER, B.F., TSOTESTI, N.G., SEBITSO, C., SIWEDI, D. and 17 WALKER, A.J., Case-control study of prostate cancer in black patients in Soweto, South Africa. Brit. J. Cancer, 65, 438–441 (1992). WEST, D.W., SLATTERY, M.L., ROBISON, L.M., FRENCH, T.K. and MAHONEY, A.W., Adult dietary intake and prostate cancer risk in Utah: a case-control study with special emphasis on aggressive tumors. Cancer Causes Control, 2, 85–94 (1991). WHITTEMORE, A.S., KOLONEL, L.N., WU, A.H., JOHN, E.M., GALLAGHER, R.P., HOWE, G.R., BURCH, J.D., HANKIN, J., DREON, D.M., WEST, D.W., TEH, C.-Z. and PAFFENBARGER, R.S., Prostate cancer in relation to diet, physical activity and body size in blacks, whites and Asians in the US and Canada. J. nat. Cancer Inst., 87, 652–661 (1995). WHITTEMORE, A.S., PAFFENBARGER, R.S., ANDERSON, K. and LEE, J.E., Early precursors of site-specific cancers in college men and women. J. nat. Cancer Inst., 74, 43–51 (1985). WYNDER, E.L., MABUCHI, K. and WHITMORE, W.F., JR., Epidemiology of cancer of the prostate. Cancer, 28, 344–350 (1971). YATANI, R., CHIGUSA, I., AKAZAKI, K., STEMMERMANN, G.N., WELSH, R.A. and CORREA, P., Geographic pathology of latent prostatic carcinoma. Int. J. Cancer, 29, 611–616 (1982). YU, H.E., HARRIS, R.E. and WYNDER, E.L., Case-control study of prostate cancer and socioeconomic factors. Prostate, 13, 317–325 (1988).