Epidemiology of Prostate Cancer

Edward Giovannucci

Elizabeth A. Platz

Lorelei Mucci

INTRODUCTION

Prostate cancer is the most commonly diagnosed cancer in American men and is among the most common cancers diagnosed in many developed countries. The classic risk factors for this cancer are older age, African American racial group, and family history of prostate cancer. Otherwise, few established modifiable risk factors have emerged, despite recent interest and intensive study of this cancer. Nonetheless, the wide variation in incidence rates among countries and increases in prostate cancer rates in groups that have migrated from countries that have low rates to those countries with high rates strongly suggest the importance of environmental factors in its etiology. A number of leads have emerged, though the evidence for each has generally not been consistent or conclusive enough to allow for definitive conclusions at this time. A strong research focus has been on nutritional and hormonal factors, and recently, an interest in inflammatory or infectious etiologies has reemerged. This chapter outlines demographic patterns of prostate cancer, major environmental and host risk factors, and the primary prevention strategies for prostate cancer.

DEMOGRAPHIC PATTERNS

Mortality and Incidence in the United States

Prostate cancer is the second most common cause of cancer death in men in the United States, accounting for over 32,020 deaths annually. The introduction of widespread screening for prostate-specific antigen (PSA) in the early to mid 1990s has caused the prostate cancer incidence rate to soar in the Unites States (1). Excluding skin cancer, prostate cancer is now the most commonly diagnosed cancer in American men by far and accounts for an estimated 217,730 new cases annually or about one third of all new cancer diagnoses in men (2). The lifetime risk of a male in the United States being diagnosed with prostate cancer is one in six (2). After decades of a slightly increasing mortality rate, beginning around 1992, the prostate cancer mortality rate has declined. The reasons for this decline remain controversial but may be attributable at least in part to earlier detection through PSA screening and subsequent treatment, and better treatment for advanced stage prostate cancer (3). Not surprisingly given the dramatically increasing incidence and the stable or reducing mortality, the 5-year survival rates increased markedly over 1974-1976 to 1983-1985 to 1992-1998 from 68% to 76% to 98%, respectively, among whites in the United States (2). Increases in survival have also been observed in other racial groups. This increased survival is likely mostly due to detection at an earlier stage of the disease and to an overdiagnosis by PSA of prostate cancers with little biologic potential to progress, although an actual improvement due to early detection and treatment may have contributed to this improvement. Overall, 5-year diseasespecific survival rates are close to 100% for men diagnosed with localized prostate cancer, but only 34% for men diagnosed with distant metastases (2).

Age

The strongest risk factor for prostate cancer is age. The characteristic logarithmic rise in the incidence rate with log age that is observed for many cancers is steepest for prostate cancer. The majority of men are diagnosed with prostate cancer at an age older than 65 years, and the vast majority of prostate cancer deaths occur in this older age group. The median age at prostate cancer diagnosis is 71 years in whites and 69 years in blacks in the United States (4). In recent years, the average age of diagnosis has dropped following the earlier detection of cancers due to the widespread use of PSA screening. However, the average age of death from prostate cancer, the late 70s, has remained relatively stable. The relatively old age at diagnosis coupled with the variable rates of progression for prostate cancer results in a number of dilemmas regarding what efforts should be put into diagnosis and treatment.

Race and Ethnicity

One of the well-recognized risk factors for prostate cancer is race. In the United States, African Americans have the highest prostate incidence rate (standardized to the 2000 United States population age standard, 1992-1999; 275.3 per 100,000 men annually) and mortality rate (75.1 per 100,000 men annually) among any racial or ethnic group. The incidence rate is 1.6-fold higher than in whites (172.9 per 100,000 men), and the disparity in mortality is even greater at 2.3-fold (32.9 per 100,000 men) times than for whites. Both incidence and mortality rates of prostate cancer for Asian/Pacific Islander, American Indian/Alaskan Native, or Hispanic are substantially lower than those for white Americans (2).

The underlying causes for this disparity in rates among races remain of great interest for both societal reasons and for potentially providing insights into the etiology of prostate cancer. Not only do African American men experience a higher incidence of this cancer, but they also appear to experience a more aggressive form of the disease. For example, among whites in the United States, 10% of prostate cancers are diagnosed with distant metastases, compared with 18% for African
American men. Reduced access to health care may contribute to the more advanced stage distribution at diagnosis among African Americans, but this factor may not account entirely for the difference. In a study conducted among men in the US military medical system, a population without economic barriers to screening, diagnosis, and treatment, the risk of being diagnosed with advanced stage prostate cancer remained higher in African American men compared with non-Hispanic white men after adjusting for socioeconomic factors (5). Moreover, in a study of male health professionals, risk of total and advanced prostate cancer remained elevated even after adjusting for known and suspected dietary and lifestyle risk factors and screening behaviors (6). Further, because these men were all highly educated health professionals, differences due to socioeconomic status were minimized.

International Patterns of Mortality and Incidence

Prostate cancer incidence rates vary dramatically among countries. However, some of the disparity in prostate cancer incidence rates between countries is likely due to differences in medical practice leading to differential rates of detection of subclinical tumors. For example, PSA screening has caused a great increase in the incidence of prostate cancer diagnosed in the United States, as well as in some other countries to a lesser degree. Prostate cancer mortality rates around the world also vary dramatically, more than 30-fold between the highest and lowest rate countries (2). Developed or “Westernized” countries have substantially higher prostate cancer rates than do developing countries (7). The lowest prostate cancer incidence and mortality rates are observed in the Far East and on the Indian subcontinent, and the highest rates occur in the Western Europe, Australia, and North America. Even adjusting for age, the mortality rate in the United States for prostate cancer is 18-fold that of China (8). Although issues such as variable attribution of cause of death and treatment may contribute to the differences in mortality rates, these factors alone are unlikely to account for such a dramatic difference in mortality rates among these countries.

Given the high rates of prostate cancer in African Americans, it is of interest to examine rates among men of African descent elsewhere in the world. Among the highest prostate cancer rates are found on Caribbean islands, including Trinidad and Tobago, which has the highest mortality rate among 45 countries evaluated (2) and Jamaica (9). In comparison, prostate cancer incidence and mortality rates in African countries such as Nigeria appear to be substantially lower than these (10). However, it remains questionable whether cancer incidence and mortality information is sufficiently comprehensive in African countries to make valid comparisons. In addition, there could be genetic differences in men from these different areas. Nonetheless, the overall evidence indicates that factors related to industrialization contribute to higher incidence and mortality rates of prostate cancer. Identifying these factors could lead to prevention strategies.

Migration Studies

Men of Asian heritage living in the United States are at lower risk for prostate cancer than white Americans but are at greater risk than men of similar ancestries living in Asia (11,12). Japanese immigrants living in Los Angeles County, California, have prostate cancer rates that are more comparable to individuals with similar ancestry but who were born in the United States than to Japanese men living in Japan (13). Moreover, these rate differences do not appear to be entirely due to differences in detection of early stage tumors between the United States and Japan (14). The changing rates of clinical prostate cancer among migrants suggest that factors affecting tumor growth or progression may vary between countries. A number of hypotheses attempting to address these rate differences have been generated.

Family History of Prostate Cancer

Another of the few clearly established risk factors for prostate cancer is a family history of prostate cancer. The relationship between family history and prostate cancer has been confirmed in population-based case-control (15,16,17,18,19,20), record linkage (21,22,23), and prospective cohort (24,25,26) studies. In general, men with either a father or a brother with the diagnosis of prostate cancer are at approximately two to three times greater risk of developing prostate cancer than are men without a family history of the disease (15,16). The magnitude of this association is modified by at least three factors. First, a man’s risk of prostate cancer risk is higher if a brother rather than father is affected compared to neither first-degree relative being affected (15,17,18,27). Second, risk is higher with increasing number of first-degree relatives with prostate cancer (for two or more affected relatives, the relative risk [RR] = 3-5) (16,18,24). Third, risk is higher if the first-degree relative was younger at diagnosis (under about 65 years, RR = 1.5-6.0) (18,21,22,24). Not surprisingly, the proportion of prostate cancer attributable to family history is greater for younger ages at diagnosis (18,22,28). With advancing years, more and more prostate cancers are considered sporadic, and fewer are due to a strong genetic component.

Family history may be a surrogate of hereditary factors, although to some extent it may reflect common environmental factors (including the in utero environment). Further support for an inherited component to prostate cancer risk is that twin studies show higher concordance for prostate cancer diagnosis between monozygotic than dizygotic twins of 18% and 3%, respectively, in a combined analysis of Swedish, Danish, and Finnish twins (29). In a similar comparison of United States veterans twins, the respective concordances were 27.1% and 7.1% (30). The observation of a higher risk if a man’s brother compared to his father is affected has been postulated to indicate that the underlying genetic inheritance of susceptibility is either X-linked or recessive (27). However, a nongenetic explanation for the higher risk if a brother is affected may result from brothers sharing a more similar environment in utero and in childhood than fathers and sons. Segregation analyses in prostate cancer kindreds support an autosomal dominant mode of inheritance (28,31,32), with the attributable risk ranging from 43% in men diagnosed under age 55 years, to 34% under age 70, and to 9% under age 85 years (28).

DIET AND NUTRITION

A western diet has been proposed as a potentially important prostate cancer risk factor, based on international comparisons of prostate cancer mortality rates and the observation that migrants from low- to high-risk geographic areas, as well as their offspring, assume the higher risk profile of their adopted countries. Many aspects of the Western diet, including energy imbalance (energy consumption exceeding energy requirements) and high consumption of fat, particularly animal fat and red meat, have been postulated as playing a role in increasing prostate cancer risk. The low rates of prostate cancer in Asia and some Mediterranean countries have pointed toward foods commonly consumed in those populations, including soy and tomato products, in preventing prostate cancer. The potential role of energy balance, composition of macronutrients, micronutrients, and phytochemicals are discussed in this section.

Energy Intake and Obesity

Energy balance describes a state of equilibrium between total energy intake versus energy expenditure. In epidemiological studies, the components describing various aspects of energy balance—total energy intake and energy expenditure are not easily measurable. Thus, measures of body mass, adiposity, and obesity have typically been used as surrogates for energy balance, or more precisely, the consequences of energy imbalance. While the etiology of obesity is complex, excess adiposity largely represents a relatively chronic state of overconsumption of energy relative to requirements. A positive or negative energy balance may influence multiple hormonal pathways that may differentially influence prostate cancer pathogenesis. Moreover, energy balance and obesity may have different effects for disease progression than for disease incidence. In the section below, we review of the literature on energy intake and obesity, as well as height because short stature may represent a surrogate of a relative energy restriction during the growth period.

In a prostate cancer animal model, an energy-restricted diet led to reduced prostate tumor angiogenesis and decreased tumor growth (33), indicating that total energy balance influences cancer progression. Yet, the epidemiological evidence regarding the effect of total energy intake on prostate cancer risk is inconclusive (34,35). In general, some studies have reported a positive association of increased total energy (36,37), while others have indicated no association of energy intake with prostate cancer incidence (38,39). In epidemiologic studies, total energy intake can be difficult to interpret, since its determinants are complex and influenced by body size, physical activity, metabolic efficiency, and energy requirements. In addition, energy intake is difficult to measure feasibly in large populations.

For obesity, the bulk of evidence suggests that excess body weight in adult life is not associated with total prostate cancer incidence (38,40,41,42,43,44,45,46,47,48,49,50,51). However, some studies have suggested a positive association between high body mass index and prostate cancer risk (52,53,54,55,56,57,58,59). A complicating factor is that in some settings, associations with body mass index may reflect lean muscle mass rather than adiposity. For example, a large Swedish study found prostate cancer risk to be more strongly associated with lean body mass than with body mass index (60). Nonetheless, overall obesity (61,62,63,64,65,66,67) and central obesity (68,69) are associated with an increased risk for men to develop advanced stage or fatal prostate cancer. Obesity could increase risk of prostate cancer mortality through direct causal mechanisms, or indirectly; for example, prostate cancer could be possibly more difficult to diagnose or treat in obese men, leading to delayed diagnosis and suboptimal therapy. Because the association between obesity and fatal prostate cancer was observed before the onset of PSA screening, the role of obesity on fatal prostate cancer risk is unlikely to be limited to a delay in diagnosis (52).

Issues related to energy balance have been especially difficult to study in earlier life periods, such as childhood and adolescence. Tallness has been used as a surrogate of high exposure to growth hormones, which are partially influenced by energy balance as a relative restriction of energy could lead to shorter stature. As has been observed for other cancers, taller height has been a risk factor in some studies of prostate cancer, particularly aggressive disease (64). In a recent study, tallness was not associated with total prostate cancer incidence, but taller men had almost a threefold higher risk of dying from prostate cancer (70). The association with height strongly suggests that factors during puberty and adolescence, possibly related to nutritional status, influence prostate carcinogenesis. For example, energy and nutritional restriction resulting in submaximal height may underlie the decreased risk of prostate cancer typically observed in populations that have not experienced economic development. The role of energy balance and imbalance throughout life is difficult to measure, but initial results indicate that they play a major role in prostate carcinogenesis and progression.

Macronutrients and Related Factors

Fat and Fatty Acid Consumption

Support for fat intake as a risk factor for prostate cancer was initiated by the observation that per capita fat consumption is strongly correlated with prostate cancer incidence and mortality rates internationally (71,72). An association between dietary fat or higher fat foods, especially animal foods such as red meat and dairy products, and prostate cancer has been further supported in several case-control studies conducted in several different populations, even after adjusting for potentially confounding factors (73). For example, in a multiethnic case-control study, an association was found between higher saturated fat intake and total and advanced prostate cancer risk separately among African Americans, whites, Chinese Americans and Japanese Americans (39). In addition, a case-control study reported positive associations for foods high in animal fat and total and advanced prostate cancer in blacks and in whites (38). In contrast, other case-control studies have shown no association for total fat or saturated fat (74,75).

The findings for dietary fat and prostate cancer from prospective cohort studies, which are less prone to recall or selection bias, are not as consistent. Many have not supported an association (76,77,78), while other studies have supported an association for total fat or fat from animal products (45), or only for advanced prostate cancer (79). A meta-analysis of seven prospective studies found no significant associations between dietary intake of total, saturated, monounsaturated, and polyunsaturated fat and prostate cancer risk (80).

Some prospective studies found stronger associations for fat intake and risk of advanced or fatal disease than for total prostate cancer (38,39,79,81). The stronger findings for advanced disease suggest that dietary fat influences late stages of carcinogenesis. Two small studies of men with prostate cancer suggest that high intake of saturated fat at the time of diagnosis is associated with an increased risk of biochemical failure (82) and prostate cancer-specific death (83). The stronger findings for advanced disease and progression, if confirmed, suggest that dietary fat may influence late stages of carcinogenesis. However, the European Prospective Investigation into Cancer and Nutrition (EPIC) study, a large European cohort study, did not find any association between total, saturated or monounsaturated fat intake and aggressive (advanced stage) prostate cancer (80). Moreover, a reduced risk of high grade prostate cancer was suggestively associated with increased intake of total fat, monounsaturated and polyunsaturated fat intake.

Studies have also examined specific fatty acid types. Some evidence, though not consistent, has suggested that α-linolenic acid, an omega-3 polyunsaturated fatty acid, increases risk of prostate cancer. This fatty acid is found in vegetable oils such as soy and canola, and less so in leafy green vegetables (84), as well as in red meat and dairy fat (79,85,86,87,88). Fatty fish such as salmon and tuna contain long chain omega-3 polyunsaturated fatty acids such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). Curiously, EPA and DHA, have been either unrelated or inversely associated with risk of prostate cancer in contrast to α-linolenic acid (see “Fish Consumption” below) (85,86,87,88,89,90,91). Because α-linolenic acid is a precursor to the longer chain omega-3 fatty acids, the positive association between α-linolenic acid and prostate cancer risk remains an enigma. Only a small proportion of α-linolenic acid is converted to long chain omega-3 fatty acids, so perhaps
other breakdown products of this fatty acid are deleterious. Alternatively, this fatty acid may be acting as a surrogate of another factor.

The essential fatty acid linoleic acid, an omega-6 fatty acid found in vegetable and soy oils, is a precursor to arachidonic acid and ultimately to the proinflammatory prostaglandins of the two series. No consistent associations between intake of (79,85) or circulating concentration of linoleic acid and prostate cancer risk have been observed (86,87). Higher concentration of erythrocyte and adipose linoleic acid, a marker of linoleic acid intake, was associated with a nonstatistically significant higher risk of prostate cancer in a small case-control study (89). Dietary intake (85) and circulating concentration (86,87) of arachidonic acid, the immediate precursor to prostaglandins, were not associated with prostate cancer in prospective studies.

Meat Consumption

It is possible that the frequent associations with prostate cancer observed with animal fat in many (though not all) studies is not related directly to fat, but to other components in animal foods. Besides being sources of fat, red meat and processed meat are sources of iron and protein, as well as added nitrites and byproducts of cooking, such as heterocyclic amines. In international studies, total meat intake is positively correlated with prostate cancer mortality rates (71,72). In some cohort studies, consumption of red meat is associated with a higher risk of prostate cancer (45,59,79,86,92), especially for processed or cured meats (92,93). A large case-control study also supported an association of red meat with advanced prostate cancer in both blacks and whites (38), but other studies are not supportive (48,76,78,80,94). In general, intake of poultry has not been a risk factor for prostate cancer in most studies (38,45,76,78,92,93), suggesting that some factor specific to red meat or processed meat is most relevant for prostate carcinogenesis.

Recently, interest has arisen in heterocyclic amines, mutagenic compounds formed during high temperature cooking of meat and fish through the condensation of amino acids with creatinine (95). In the grilling of red meat, heterocyclic amine carcinogens form, including 2-amino-1-methyl-6-phenylimidazo(4,5-β)pyridine, which causes prostate cancer when fed to rats (96). A case-control study reported no consistent associations for estimated intake of total or specific heterocyclic amines, although a 1.7-times higher risk of prostate cancer was noted for those who consume well-done beef steak that was fried/grilled, or barbequed compared to noneaters/other cooking methods (97). However, a case-control study did not show an association for roast or grilled meat (74). In the Agricultural Health Study, with detailed information on cooking patterns, there was no association with prostate cancer specifically for pan-frying or grilling meats. However, higher intake of well-done meats was associated with 30% greater risk of prostate cancer, and even stronger for advanced disease (98).

Fish Consumption

Populations with a high consumption of fish, such as Japan and among Eskimos in Alaska, have lower rates of prostate cancer than populations with Western food habits (99,100,101). Fish contain the long-chain marine omega-3 polyunsaturated fatty acids EPA and DHA, which may lower prostate cancer risk and progression (102). The findings from epidemiologic studies of fish intake and prostate cancer risk and progression have reported inconsistent findings, which may be attributed to the types of fish consumed in different populations as well as the prostate cancer outcome measured. Of four case-control studies, three (103,104) reported associations between high intake of fish and lower incidence of prostate cancer. Findings from prospective studies are more mixed (45,74,77,91,94,105,106), with only 2 of 10 reporting significant inverse associations for total prostate cancer incidence (91,106). In a Swedish twin cohort (91), where men traditionally consume fatty fish such as salmon and herring, moderate consumption fish was associated with a lower risk of prostate cancer. In the US Health Professionals Follow-up Study (HPFS), consuming fish three or more times per week was associated with a reduced risk of advanced stage or metastatic prostate cancer (106). The Swedish twin study (91), as well the US Physicians’ Health Study (107), and a Japanese cohort of men (108) found higher baseline intake of fish was associated with a reduced risk of cancer-specific mortality among men diagnosed with prostate cancer. Moreover, in the HPFS, regular consumption of fish after prostate cancer diagnosis was associated with a 50% reduction in disease progression (109). The association between intake of fish and reduced prostate cancer progression is intriguing, although not confirmed in all studies (94). For example, a Japanese study found that high fish consumption was associated with an increased risk of prostate cancer (110).

Dairy Products and Calcium

Higher consumption of dairy products has been associated with a higher risk of prostate cancer in various types of epidemiologic studies. Countries with greater per capita consumption of milk tend to have higher prostate cancer mortality rates (71,72,111). High intake of dairy products has been associated with an increased risk of prostate cancer in several case-control (112,113,114,115) and cohort studies (45,113,116,117,118,119). Positive associations have been observed for total dairy intake, as well as specifically for higher intake of milk (113,119), cheese (120), and yogurt (119). A meta-analysis of 11 cohort studies reported increases or suggestive increases in risk associated with intake of total dairy, milk, and cheese (121). Most (94,118,119,120,122) but not all (123,124) studies published since this meta-analysis have tended to support an association between higher milk or dairy consumption and prostate cancer risk.

Dairy products are common dietary sources of calcium and animal fat, and in some settings dairy products are supplemented with vitamin D. The correlation between dairy foods and these nutrients create challenges in trying to disentangle the independent effects of dairy, calcium and vitamin D. Cohort studies that have tried to parse out effects suggest calcium may be the predominant player in explaining positive associations with prostate cancer. In studies that simultaneously consider dairy intake and calcium, RR estimates for dairy are attenuated compared to calcium (113,118,120). In a recent analysis of the EPIC cohort, dairy protein and dairy calcium were both similarly associated with risk of prostate cancer (94).

Whether calcium specifically increases risk of prostate cancer has been evaluated in several case-control (112,114) and cohort (94,113,115,117,118,125) studies, with positive associations reported for total or advanced prostate cancer risk. Moreover, very high intake of calcium through diet or supplementation was associated with significant excess risks (113,115,125). Several studies have reported stronger associations between high intake of calcium and risks of aggressive forms of prostate cancer, defined by high-grade (70), or advanced or lethal prostate cancer (113,125). In the National Health and Nutrition Examination Survey (NHANES) study, higher circulating levels of calcium were associated with a suggestive increased risk of total cancer incidence and an increased risk of fatal disease (126), although the number of prostate cancer events was small. Moreover, plasma calcium levels are tightly regulated and it is unclear how this finding relates to dietary calcium intake. Not all studies have confirmed a positive association for calcium (38,93,127,128), while a randomized trial of calcium supplements and colorectal adenoma
(129) found no positive association with prostate cancer as secondary endpoints. However, the follow-up time was short and the cases were primarily PSA-detected, early stage cancers. Although the association between calcium intake and prostate cancer risk remains controversial, this hypothesis requires resolving because many men in the United States take calcium supplements with expectations of health benefits.

Fruits, Vegetables, Legumes, and Micronutrients

Fruits, Vegetables, and Legumes Including Soy Products

Fruits and vegetables are known to possess a wide spectrum of phytochemicals; some of these compounds may protect against cancer in general or prostate cancer specifically. The relevant compounds include antioxidants, which have received most interest; other vitamins and essential minerals; and nonessential, but bioactive compounds. Total fruit and vegetable intakes, which have been related to lower risk of some cancers, have generally not been associated with lower prostate cancer risk in most studies (20,38,45,48,76,77,78,130,131), but some exceptions have suggested benefit (132,133). Overall, the evidence to date does not suggest that total fruits and vegetables will be strongly associated with a reduced risk of prostate cancer.

Although total fruit and vegetable consumption may incorporate a wide spectrum of potentially beneficial compounds, examining total fruits and vegetables may be too simplistic an approach if there is some benefit from a specific subgroup of these items. Three botanical families or groups of vegetables are of special interest. Inverse associations are supported by some studies for tomato-based foods, Brassica vegetables, and soy and other legumes, and allium vegetables such as garlic and onions (74,131,134). Each of these groups has major characterizing phytochemicals, with anticancer properties. Thus, each is discussed further below.

Tomatoes

Tomatoes are rich in lycopene, a carotenoid with well-documented antioxidant effects. A hypothesized mechanism for cancer prevention by tomatoes or lycopene is via reduction in cellular oxidative stress (135), which can cause chronic inflammation and might therefore be related to prostate carcinogenesis. In addition, tumor cells produce 10 times the levels of reactive oxygen species compared to normal cells (136), leading potentially to prostate cancer progression. Antioxidants may lower risk, particularly for advanced prostate cancer by quenching free radicals and thus ameliorating damage from consequences of chronic inflammation, and also by downregulating tumor angiogenesis.

A summary of epidemiological findings on tomatoes and prostate cancer risk through 2003 were examined in a metaanalysis (137), including results from 11 case-control and 10 prospective cohort or nested case-control studies of circulating lycopene levels. Included were studies that presented data on the intake of tomatoes, tomato products, or dietary lycopene. Compared with men with low intake of tomato products (first quantile of intake), consumers of higher amounts of raw tomato (fifth quantile of intake) had an 11% lower risk of prostate cancer. Cooked or processed tomato products, such as tomato sauce, tomato soup, and ketchup, offer more readily bioavailable sources of lycopene than fresh tomatoes (138). Accordingly, some epidemiologic studies have found stronger inverse associations for tomato sauce while reporting weaker results for raw tomato intake (139). The meta-analysis suggested a 19% risk reduction associated with higher intakes of cooked tomato products (137). For cohort studies conducted prior to PSA screening, only a study based in the Netherlands (131) found no appreciable association between tomato consumption and prostate cancer risk. However, tomato consumption appeared to be about 10-fold lower than that in US-based studies.

The majority of serum or plasma-based studies of lycopene have also found inverse associations for high lycopene levels (137). Serum levels of lycopene were not associated with prostate cancer risk in a Japanese American population in Hawaii (140). However, serum lycopene levels were quite low in that population, with median concentration among controls of 134 ng/mL, which was 60% to 80% lower than those among men in studies of mainly whites in the United States (141,142,143). However, another recent study conducted in the United States, with relatively high levels of lycopene, showed no association between lycopene and risk of total prostate cancer (144). This study included 692 incident prostate cancer cases, although the number of advanced stage cases of prostate cancer was limited as the study was conducted in a population with extensive PSA screening.

Tomatoes and lycopene are also interesting to study with respect to prostate cancer progression. Epidemiological studies generally point to a stronger reduction in risk of advanced stage or lethal prostate cancer. For example, in the HPFS, the associations comparing high and low consumption of tomato sauce were higher for advanced stage than for total prostate cancer (70). Moreover, the association was stronger for cancers diagnosed prior to the introduction of PSA screening than for later diagnosed cancers. Also in this cohort, higher tomato sauce consumption after cancer diagnosis was associated with a lower risk of disease recurrence or progression (109). The idea that lycopene or tomatoes would be associated more strongly with prostate cancer progression was also supported in an analysis from the EPIC study (145) based on 966 total cases and 205 advanced stage cases of prostate cancer. While no association was seen for total prostate cancer, men in the top quintile of plasma lycopene had a significantly reduced risk of advanced stage prostate cancer.

Cruciferous Vegetables

Cruciferous vegetables such as broccoli, cauliflower, and brussel sprouts are rich in glucosinolates. Isothiocyanates and sulforaphane, the hydrolytic products of glucosinolates, have been shown to upregulate apoptosis, downregulate metastasis and angiogenesis, and suppress inflammatory NF-kappa B pathways in prostate cancer experimental models (146). Moreover, sulforaphanes induce phase II xenobiotic metabolizing enzymes that protect cells from DNA damage (147). Substantial experimental data suggest a benefit of high intake of cruciferous vegetables on prostate cancer (148,149,150).

Case-control studies have generally reported inverse associations between high consumption of broccoli or total cruciferous vegetables and overall prostate cancer (133,151,152), but results from prospective studies have been less consistent. In the screening arm of the PLCO screening trial, higher intake of cruciferous vegetables was associated with a reduced risk of total prostate cancer (153), with similar associations for broccoli and cauliflower. The inverse association with cruciferous vegetables was also seen in cohorts of men in the United States (154) and Netherlands (131) but not in three others studies (141,155,156). In the HPFS, only a weak nonsignificant inverse association was found, but this inverse association was strengthened and became significant when long-term intake of cruciferous vegetables was considered (154). Among three studies that specifically evaluated cruciferous intake in relation to extraprostatic prostate cancer (133,153,154), two found stronger associations for advanced stage disease (133,153), and the other (154) did not.

Only gold members can continue reading. Log In or Register to continue