- Email: [email protected]
Comparisons of prostate cancer mortality rates with dietary practices in the United States
Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
Original article
Comparisons of prostate cancer mortality rates with dietary practices in the United States Janet Laura Colli, M.D.*, Albert Colli, B.S.Ch.E. Division of Urology, Department of Surgery, University of Alabama at Birmingham, Birmingham, AL, USA Received 25 November 2004; received in revised form 21 February 2005; accepted 3 March 2005
Abstract From 1930 to 1992, prostate cancer mortality rates in the United States doubled and then declined somewhat until 2000. The objective of this study is to determine whether variations in prostate cancer mortality rates correlate with dietary changes that occurred over that period. Simple linear regression models were applied to age-adjusted prostate cancer mortality rates and per-capita consumption rates for 18 foods from 1930 to 2000. Correlation coefficients were calculated while comparing food consumption rates to prostate cancer mortality rates for the same year. Correlation coefficients were then recalculated when the prostate cancer mortality rates were compared with food consumption rates that occurred: 1 yr; 2 yr; 3 yr; and continuing in progression for 21 yr before the occurrence of the prostate cancer mortality. The largest positive correlation coefficients were associated with the consumption of: total meat (red meat, poultry and fish) (R ⫽ 0.83, T between 0 and 1); added fats and oils (R ⫽ 0.83, T ⫽ 21); ice cream (R ⫽ 0.83, T ⫽ 20); margarine (R ⫽ 0.81, T ⫽ 4); salad/cooking oil (R ⫽ 0.82, T between 3 and 4) and; vegetable shortening (R ⫽ 0.81, T between 1 and 2) where R is the correlation coefficient and T is the time in years between consumption and mortality. In conclusion, this study found strong positive correlations between prostate cancer mortality and the consumption of: total meat; added fats and oils, ice cream, salad/cooking oils, margarine, and vegetable shortening. The connection between total meat consumption and prostate cancer risk is consistent with previous studies in the literature. The link between salad/cooking oil consumption and prostate cancer risk may be consistent with past studies which suggest that -linolenic acid (a component of salad/cooking oils) is a suspected risk factor for prostate cancer. © 2005 Elsevier Inc. All rights reserved. Keywords: Prostate cancer; Foods; Diet; Epidemiology
1. Introduction In the period from 1930 to 2000 prostate cancer mortality rates in the United States varied considerably. Over that period, there have been changes in dietary practices as people acquired different tastes and consumed foods that became commercially available. The purpose of this study is to compare age-adjusted prostate cancer mortality to the per-capita consumption of various foods to determine if there are correlations that suggest potential associations. Fortunately, average per-capita consumption levels for various foods, based on food disappearance data, have been compiled by the United States Department of Agriculture since 1909 [1]. More recently, a study from an American Cancer Society publication presented age-adjusted cancer
* Corresponding author. Tel.: ⫹1-214-707-8801; fax: ⫹1-205-9341470. E-mail address: [email protected] (J.L. Colli). 1078-1439/05/$ – see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.urolonc.2005.03.020
mortality for cancers at various sites (including prostate cancer) for the period from 1930 to 2000 [2]. The existence of these two data bases allowed this study to be conducted. In this study, we selected prostate cancer mortality rather than incidence as the focus of the analysis for several reasons. Incidence data are only available for a much more limited time period [3] than mortality data reducing the robustness of the analysis. In addition, prostate-specific antigen testing introduced in the late 1980s increased the number of prostate cancer incidences detected and would have been a confounding factor to the dietary factors analyzed in this study. Although there has been a decline in prostate cancer mortality rates since the early 1990s, this decrease does not appear to be associated with early detection from prostate-specific antigen testing and treatment [4]. There are other reasons other than dietary changes that may have caused changes in prostate cancer mortality over that 71 yr period: toxic contaminants have been introduced into the environment because of industrialization; ethnic
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398 Table 1 Cross correlation coefficients between prostate cancer mortality rates and per-capita consumption rates of animal products in the United States from 1930 to 2000 Time* (years)
Total meat
Red meat
Beef
Pork
Chicken
Fish
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
0.86 0.86 0.85 0.85 0.84 0.82 0.81 0.79 0.79 0.79 0.78 0.78 0.77 0.77 0.77 0.76 0.76 0.74 0.73 0.71 0.70 0.69
0.67 0.69 0.7 0.71 0.72 0.72 0.72 0.71 0.72 0.73 0.74 0.74 0.74 0.75 0.75 0.75 0.75 0.74 0.72 0.70 0.69 0.67
0.68 0.70 0.72 0.73 0.74 0.74 0.75 0.75 0.75 0.76 0.76 0.76 0.75 0.75 0.74 0.74 0.74 0.73 0.72 0.70 0.68 0.64
0.5 0.48 0.45 0.43 0.44 0.43 0.38 0.34 0.35 0.37 0.39 0.39 0.40 0.43 0.41 0.41 0.41 0.42 0.41 0.39 0.42 0.45
0.79 0.78 0.78 0.77 0.76 0.75 0.73 0.73 0.73 0.73 0.74 0.74 0.73 0.73 0.72 0.70 0.69 0.67 0.65 0.65 0.64 0.64
0.78 0.75 0.71 0.67 0.63 0.60 0.57 0.55 0.50 0.48 0.42 0.37 0.35 0.32 0.30 0.28 0.25 0.24 0.22 0.20 0.16 0.10
* Time in years from consumption to mortality.
changes in the population have occurred from migration; pesticide use has increased and; hormones have been introduced into the food supply to increase the growth of animals for human consumption. Nevertheless, examining the relationship between prostate cancer mortality and dietary practices over such a long period of time as 71 yr may provide some insight into the causes of prostate cancer.
2. Methods 2.1. Data sources Age-standardized prostate cancer mortality rates from 1930 to 2000 were obtained from Jemal et al. [2] published by the American Cancer Society. The authors of the study obtained mortality data from 1930 to 2001 from the National Center for Health Statistics [5]. In their study, causes of death were coded and classified according to the International Classification of Diseases (ICD-8, ICD-9, and ICD10) [6 – 8]. The paper presented age-adjusted mortality rates that were standardized to the 2000 United States standard population and expressed per 100,000 population [2]. Per-capita food consumption rates from 1920 to 2000 were obtained from the food consumption data system [1] compiled by the Economic Research Service (ERS) for the United States Department of Agriculture. Information on domestic production, imports, exports and losses were used
391
by ERS statisticians in estimating average per-capita consumption levels for foods in the database. ERSs food availability data represent the food supply, or the disappearance of food into the food marketing system, because they are normally calculated as the residual of a commodity’s total annual available supply after subtracting measurable uses, such as farm inputs (feed and seed), exports, ending stocks, and industrial uses. For this study we selected per-capita consumption rates for the following foods, or food categories for which data was available for the period from 1920 to 2000: total meat (red meat, poultry, fish), red meat, beef, pork, chicken, fish, dairy, milk, cheese, ice cream, eggs, flour and cereal, fats and oils, butter, margarine, salad/ cooking oil, shortening (animal fat and vegetable shortening) and vegetable shortening. In all cases except eggs, the units of the food consumption rates are provided in kilograms per year. The units for eggs are: whole eggs per year. 2.2. Statistical analysis Simple linear regression models were applied to prostate cancer mortality rates and per-capita consumption rates for eighteen foods (listed above) for the period from 1930 to 2000. Correlation coefficients were calculated when food consumption rates were compared to prostate cancer mortality rates for the same year. Correlation coefficients were then recalculated when the prostate cancer mortality rates were compared with food consumption rates that occurred:
Table 2 Cross correlation coefficients between prostate cancer mortality rates and per-capita consumption rates of dairy products, eggs and flour in the United States from 1930 to 2000 Time* (years)
Dairy
Milk
Cheese
Ice cream
Eggs
Flour
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
⫺0.49 ⫺0.46 ⫺0.43 ⫺0.41 ⫺0.40 ⫺0.41 ⫺0.40 ⫺0.39 ⫺0.37 ⫺0.33 ⫺0.30 ⫺0.24 ⫺0.22 ⫺0.16 ⫺0.09 ⫺0.02 0.03 0.06 0.09 0.14 0.18 0.20
⫺0.59 ⫺0.56 ⫺0.54 ⫺0.53 ⫺0.52 ⫺0.53 ⫺0.53 ⫺0.52 ⫺0.51 ⫺0.48 ⫺0.47 ⫺0.42 ⫺0.41 ⫺0.36 ⫺0.29 ⫺0.22 ⫺0.18 ⫺0.16 ⫺0.16 ⫺0.11 ⫺0.08 ⫺0.07
0.79 0.79 0.79 0.78 0.78 0.77 0.76 0.76 0.75 0.74 0.73 0.72 0.71 0.71 0.71 0.71 0.71 0.70 0.71 0.71 0.70 0.69
0.67 0.68 0.68 0.67 0.67 0.67 0.68 0.68 0.69 0.70 0.71 0.72 0.72 0.73 0.74 0.76 0.78 0.80 0.81 0.83 0.83 0.83
⫺0.50 ⫺0.50 ⫺0.50 ⫺0.50 ⫺0.51 ⫺0.50 ⫺0.50 ⫺0.48 ⫺0.45 ⫺0.40 ⫺0.36 ⫺0.33 ⫺0.28 ⫺0.23 ⫺0.20 ⫺0.16 ⫺0.11 ⫺0.06 ⫺0.02 ⫺0.01 0.03 0.09
⫺0.45 ⫺0.51 ⫺0.56 ⫺0.61 ⫺0.66 ⫺0.70 ⫺0.74 ⫺0.78 ⫺0.80 ⫺0.81 ⫺0.82 ⫺0.84 ⫺0.85 ⫺0.87 ⫺0.88 ⫺0.89 ⫺0.89 ⫺0.90 ⫺0.90 ⫺0.91 ⫺0.91 ⫺0.91
* Time in years from consumption to mortality.
392
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
Table 3 Cross correlation coefficients between prostate cancer mortality rates and per-capita consumption rates for added fats in the United States from 1930 to 2000 Time* (years)
Fats/oils
Butter
Margarine
Salad and cooking oil
Shortening
Vegetable shortening
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
0.72 0.73 0.72 0.71 0.71 0.70 0.70 0.70 0.72 0.74 0.75 0.74 0.73 0.73 0.73 0.74 0.75 0.77 0.79 0.81 0.82 0.83
⫺0.87 ⫺0.87 ⫺0.86 ⫺0.86 ⫺0.85 ⫺0.84 ⫺0.83 ⫺0.82 ⫺0.82 ⫺0.81 ⫺0.80 ⫺0.78 ⫺0.76 ⫺0.75 ⫺0.75 ⫺0.75 ⫺0.75 ⫺0.75 ⫺0.74 ⫺0.75 ⫺0.76 ⫺0.76
0.80 0.80 0.80 0.81 0.81 0.81 0.80 0.80 0.80 0.81 0.80 0.79 0.79 0.78 0.78 0.79 0.79 0.79 0.80 0.80 0.80 0.80
0.81 0.82 0.82 0.82 0.82 0.81 0.81 0.81 0.80 0.80 0.81 0.81 0.80 0.81 0.81 0.81 0.80 0.80 0.81 0.81 0.80 0.79
0.34 0.31 0.28 0.24 0.22 0.19 0.17 0.16 0.17 0.21 0.20 0.11 0.02 ⫺0.03 ⫺0.07 ⫺0.12 ⫺0.16 ⫺0.15 ⫺0.09 ⫺0.14 0.25 0.36
0.80 0.81 0.81 0.80 0.78 0.77 0.77 0.77 0.79 0.79 0.79 0.76 0.74 0.71 0.69 0.67 0.67 0.68 0.69 0.70 0.70 0.70
* Time in years from consumption to mortality.
1 yr; 2 yr; 3 yr; and continuing in progression for 21 yr before occurrence of the prostate cancer mortality. Cross correlation coefficients of the independent variables were calculated to identify collinearity (linear or nearly linear relationships) between the consumption variables that exhibited a strong correlation with prostate cancer mortality. Collinearity between the independent variables causes problems for step-wise multiple regression analysis and makes it difficult to make any conclusions from the results. All the data in this study was analyzed using the software program SPSS® Base for Windows®, version 12.0, by SPSS, Inc. (Chicago, IL).
3. Results Correlation coefficients from regressing prostate cancer mortality against independent variables (foods, or food categories) are shown in Tables 1 through 3. The largest positive correlation coefficients were associated with the consumption of: total meat (red meat, poultry and fish) from zero to 1 yr before the prostate cancer mortality (R ⫽ 0.86); added fats and oils 21 yr before mortality (R ⫽ 0.83); ice cream 21 yr before mortality (R ⫽ 0.830); margarine 4 yr before mortality (R ⫽ 0.81); salad and cooking oil from 2 to 3 yr before mortality (R ⫽ 0.82); and vegetable shortening from 1 to 2 yr before mortality (R ⫽ 0.81) where R is the correlation coefficient. The largest negative correlation coefficients were associated with the consumption of flour 19
yr before mortality (R ⫽ ⫺0.91); and butter from zero to 1 yr before mortality (R ⫽ ⫺0.87). Neither red meat (R ⫽ 0.67), poultry (R ⫽ 0.80) or fish (R ⫽ 0.78) individually had a correlation coefficient that was as high as the total meat category. U.S. per-capita total meat consumption rates and age-adjusted prostate cancer mortality rates for the period 1930 to 2000 are shown in Fig. 1. Total meat consumption rates versus age-adjusted prostate cancer mortality rates for the period 1930 to 2000 are shown in Fig. 2. On this graph, each point corresponds to an annual prostate cancer mortality rate mated with the total meat consumption rate that occurred the previous year. Salad/cooking oil and vegetable oil had the next highest positive correlation coefficient of the independent variables investigated. U.S. per-capita salad/cooking oil and vegetable oil consumption rates together with age-adjusted prostate cancer mortality rates for the period 1930 to 2000 are shown in Fig. 3. Salad/cooking oil consumption rates versus age-adjusted prostate cancer mortality rates for the period 1930 to 2000 are shown in Fig. 4. On this graph, each point corresponds to an annual prostate cancer mortality rate mated with a salad/cooking oil consumption rate that occurred 3 yr prior. Vegetable shortening and margarine both had high positive correlation coefficients with prostate cancer mortality whereas butter had a strong negative correlation coefficient. U.S. per-capita vegetable shortening, margarine and butter consumption rates together with age-adjusted prostate can-
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
393
between the consumption variables. The correlation matrix from this analysis generated the following correlation coefficients between the independent variables: total meat, salad/ cooking oil (R ⫽ 0.95); total meat, butter (R ⫽ ⫺0.94); margarine, butter (R ⫽ ⫺0.96); salad/cooking oil, shortening (R ⫽ 0.96); salad/cooking oil, vegetable shortening (R ⫽ 0.97). These results indicate that some of the independent variables that correlated closely with prostate cancer mortality were collinear and highly interconnected. Collinearity between the independent variables makes it difficult to make any conclusions from the results of stepwise multiple regression analysis. As a result, a step-wise multiple regression analysis was not performed to develop a mathematical model for the relationship between the independent variables and prostate cancer mortality.
4. Discussion
Fig. 1. U.S. prostate cancer mortality and per-capita annual total meat consumption (Kg/y).
cer mortality rates for the period 1930 to 2000 are shown in Fig. 5. Cross correlation coefficients of the independent variables were calculated to determine the degree of collinearity
In this study, the strongest dietary factors that correlated with increased prostate cancer mortality were the consumption of: total meat (that includes red meat, poultry, and fish); and certain vegetable oils categories (total added fats and oils, margarine, salad/cooking oils, and vegetable shortening). The category “total meat” results in the consumption of animal fats that are contained in red meat, poultry and fish. Ecological correlation studies in the 1970s showed a positive association between saturated fat consumption and prostate cancer incidence or mortality [9 –11]. Armstrong and Doll [9] hypothesized that dietary fat may be a major cause of prostate cancer based on a correlation coefficient of 0.74 between national per-capita consumption fat and pros-
Fig. 2. U.S. per-capita total meat consumption versus prostate cancer mortality (1930 –2000).
394
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
Fig. 3. U.S. prostate cancer mortality and per-capita salad/cooking oil consumption (Kg/y).
tate cancer mortality. More recently, an international study [12] comparing per-capita food consumption rates and prostate cancer mortality rates from different countries reported correlation coefficients of 0.78 for animal fat. In a similar international study [13], the correlation coefficient between the consumption of animal fats and prostate cancer risk were found to be 0.61. Many case-control studies have examined the association between animal fat and prostate cancer [14 –26]. Although the 13 studies differed in terms of study design, only four studies [19,22,25,26] failed to show a positive association with total fat intake. The association between fat intake and prostate cancer has been investigated in six cohort studies [27–32]. A positive association was reported in four of the studies [27–30] while two other studies did not detect an association [31,32]. Dietary fat intake from animal sources has been more consistently linked to prostate cancer than any other food category. These past studies, taken in totality, support an association between animal fats and prostate cancer risk. The high correlation coefficient between prostate cancer risk and fish consumption found in the current study runs counter to some of the evidence in the literature. Although the consumption of saturated fats have been associated with
increased prostate cancer risk, experimental studies suggest that marine fatty acids, particularly the long-chained eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) inhibit the proliferation of prostate cancer cells in vitro and reduce the progression of these tumors in laboratory animals [33]. Whether the consumption of marine fatty acids from fish can reduce prostate cancer risk in humans is unclear. In a recent prospective study of 47,882 men, eating fish more than three times per week was associated with a reduced risk of prostate cancer, and the strongest association was with metastatic cancer [34]. In a population-based cohort study of 6,272 Swedish men with 30 yr of follow-up, men who ate no fish had a two to three fold increase of prostate cancer than those who ate moderate or high amounts [35]. However, a cohort study of prostate cancer in Adventist men showed several trends of increasing risk of prostate cancer mortality with the consumption of fish [27]. In a case-control study of blood donors to a serum data bank who later developed prostate cancer, and a matched set without prostate cancer, there was no clear association between risk of prostate cancer and the presence of EPA and DHA fatty acids in the blood [36]. In another case-control study, reduced prostate cancer risk was associated with high levels of erythrocyte biomarkers for EPA and DHA [37]. In two international studies, fish consumption rates from many countries were compared to prostate cancer mortality rates to determine whether there is a correlation. In one study, the correlation coefficient for the period from 1961 to 1990 was 0.015 indicating almost no association [13]. In the other study, the correlation coefficient was ⫺0.39 suggesting that fish consumption may reduce prostate cancer risk [12]. A recent review of the literature concluded that current epidemiological evidence provides little support that marine fatty acids provide protection from prostate cancer risk [38]. Although existing epidemiological studies regarding the consumption of marine fatty acids and prostate cancer risk are inconclusive, there is little evidence that consumption of fish is associated with increased prostate cancer risk as suggested by the current study. However, reports of toxic contamination of domestic fish and shellfish are a potential cause for an association between fish consumption and increased prostate cancer risk. From 1992 to 1998, fish samples were collected from 233 stream sites by the U.S. Geological Survey’s National Water Quality program. Tissue composites from whole fish were analyzed for mercury, polychlorinated biphenols (PCBs), chlorodane, dioxane (2,3,7,8-tetrachlorodibenzo-p-dioxin) and dichlorodiphenyltrichlorethane (DDT), and trace elements. More than 90% of the sampled fish had at least one contaminant and about half had at least five contaminants [39]. The U.S. Environmental Protection Agency’s National Listing of Fish and Wildlife Advisories listed 2,838 advisories that were issued by state and local officials to protect public health in 2000; mercury, PCBs, chlordane, dioxane and DDT were responsible for 99% of the consumption advisories [39]. The Department of Health and Human Services
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
395
Fig. 4. U.S. per-capita salad and cooking oil consumption versus prostate cancer mortality (1930 –2000).
has concluded that PCBs, dioxane and DDT may reasonably be anticipated to be carcinogens. One study has implicated organochlorine pesticides, similar to those found as contaminants in fish and shellfish, as the cause for the increased rate of prostate cancer among farm workers [40]. The National Academy of Science’s committee to review the effects of exposure to herbicides in Vietnam veterans concluded that there was limited suggestive evidence linking herbicide exposure to prostate cancer [41]. In this current study, the consumption of salad/cooking oils correlated with increased prostate cancer mortality. Salad/cooking oils are a major source of -linolenic acid (an essential polyunsaturated fatty acid) in the U.S. diet. Giovannucci et al. [29] and Gann et al. [42] found: a strong positive association between -linolenic acid and prostate cancer risk; no clear linear relation across quartiles of exposure suggesting a threshold effect; that low levels of linoleic acid, and another polyunsaturated fatty acid, may further exaggerate the effect. More recently, in a cohort study of 47,866 men, Leitzmann et al. [43] found that -linolenic acid was unrelated to total prostate cancer risk but strongly related to advanced prostate cancer risk. Several case-controls studies [44,45] suggest an association with the consumption of -linolenic acid and prostate cancer risk. The author of a recent review of the literature [38] and a meta-analysis [46] concluded that -linolenic acid may be a risk factor for prostate cancer. From 1930 to 2000, salad/cooking oil consumption in the U.S. increased from 6.2 to 35.2 pounds per year resulting in an increase in per-capita -linolenic acid intake. It has been estimated that per-capita -linolenic acid consumption in the U.S. was 1.4 grams per day in 1990 [47]. This is based primarily on the consumption of soybean and canola oil,
which contain high levels of -linolenic acid: 7.4% and 9.8%, respectively [47]. Per-capita -linolenic acid consumption from vegetable oils is estimated to be about 0.1 grams per day in 1930, or about 10% of the level at the end of the 20th century. The margarine and vegetable shortening categories, both of which contain low levels of -linolenic acid [47] but high levels of hydrogenated polyunsaturated fatty acids or trans fatty acids [48], correlate strongly with increasing prostate cancer mortality. No information was found in the literature on the relationship between trans fatty acid consumption and prostate cancer risk although the National Academy of Science has recommended minimizing the consumption of trans fatty acids since they provide no benefit to human health [48]. Shortening, but not margarine, is often thermally stressed because it is used to fry foods at elevated temperatures. Thermally stressing polyunsaturated fatty acids found in vegetable shortening during routine frying or cooking, generates high levels of oxidized dietary fat containing both cytotoxic aldehydic products [49] and monoepoxy fatty acids [50]. It should be noted that there was a high correlation coefficient between prostate cancer mortality and vegetable shortening but not shortening that also contain saturated animal fats. Oxidized dietary fats have been reported to exert teratogenic actions on experimental animals [51], colon carcinogenicity in rodents [52], suppress gene expression of lipogenic enzymes in the liver of rats [53,54] and generate components that exert DNA-cleaving activity [55]. No information was found in the literature regarding the relationship between thermally stressed polyunsaturated fatty acids and prostate cancer risk. Butter was found to correlate negatively with prostate
396
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
Fig. 5. U.S. prostate cancer mortality rate and vegetable shortening, margarine or butter consumption.
cancer risk. However, it is not clear that this correlation results from protections from prostate cancer risk provided by butter consumption or by decreased risk from reduced margarine consumption. Butter and margarine are strongly negatively interrelated; the correlation coefficient between the two independent variables is ⫺0.96. The strong negative correlation between prostate cancer mortality and flour consumption found in this study is consistent with the results of two international studies [12,13] comparing per-capita food consumption rates and prostate cancer mortality rates from different countries. The maximum correlation coefficient occurred when the time period between consumption and prostate cancer mortality was: from zero to 1 yr for total meat; 21 yr for added fats and oils; 20 yr for ice cream; 4 yr for margarine; from 2 to 3 yr for salad/cooking oil; and from 1 to 2 yr for vegetable shortening. This would suggest that if the association was real and not coincidental, these time periods would represent the latency period between exposure and prostate cancer mortality. There is a high degree of uncertainty in these time estimates since it is not clear that small
changes in the value of the regression coefficient (which established the peak correlation in many cases) are significant. The correlation coefficients between prostate cancer mortality and several food categories identified in this study are very high: 0.94 between total meat and salad/cooking oils; and 0.97 between vegetable shortening and salad/cooking oils. This means that many of the independent variables that correlate strongly with prostate cancer mortality are highly interdependent suggesting that only some of them may actually affect prostate cancer risk. This study has many sources of uncertainty. For example, per-capita food intake levels are obtained from food that is produced and imported, reduced by food that is exported and wasted. It does not account completely for food wastage, or foods fed to pets although animal feed and seed usage on farms is taken into consideration. Also, food disappearance levels represent averages for the population and do not take into consideration variations that can occur throughout the country because of affluence or ethnicity. More importantly, differences in food consumption patterns between men having advanced prostate cancer and the averages for the population (as used in this study) are not known. Since the results of this study have a high degree of uncertainty, they should not be used for dietary recommendations but may provide additional insights into risk factors for prostate cancer that need to be corroborated or disproved by further research. In particular, the results from this study suggest further research into the relationship between prostate cancer risk and the consumption of: -linolenic acid; trans fatty acids and; thermally stressed polyunsaturated fatty acids.
References [1] Economic Research Service, United States Department of Agriculture. Food consumption (per capita) data system. Available at: http:// www.ers.usda.gov./Data/FoodConsumption/. Accessed July 2004. [2] Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. Cancer J Clin 2004;54:8 –29. [3] Surveillance, Epidemiology, and End Results (SEER) Program (http://www.seer.cancer.gov/) SEER*Stat Database: Incidence SEER 9 Regs Public-Use, Nov 2002 Sub (1973–2000), National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch. [4] Brenner H, Arndt V. Long-term survival rates of patients with prostate cancer-specific antigen screening era: population-based estimates for the year 2000 by period analysis. J Clin Oncol 2005;23:441–7. [5] National Center for Health Statistics, Division of Vital Statistics, Centers for Disease Control. Available at: http://www.cdc.gov/nchs/ nvss.htm. Accessed July 2004. [6] Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death. Vol. 1, 10th revision. Geneva: World Health Organization, 1992. [7] Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death. Vol. 1, 9th revision. Geneva: World Health Organization, 1975.
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398 [8] Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death. Vol. 1, 8th revision. Geneva: World Health Organization, 1967. [9] Armstrong B, Doll R. Environmental factors and cancer incidence and mortality in different countries, with special references to dietary practices. Int J Cancer 1975;15:617–31. [10] Howell MA. Factor analysis of international cancer mortality data and per capita food consumption. Br J Cancer 1974;29:328 –36. [11] Blair A, Fraumeni JF Jr. Geographic patterns of prostate cancer in the United States. J Natl Cancer Inst 1978;61:379 – 84. [12] Hebert JR, Hurley TG, Olendski BC, Teas J, Ma Y, Hampl JS. Nutritional and Socioeconomic factors in relation to prostate cancer mortality: a cross national study. J Natl Cancer Inst 1998;90:1637– 47. [13] Ganmaa D, Li XM, Wang J, Qin LQ, Wang PY, Sato A. Incidence and mortality of testicular and prostatic cancers in relation to world dietary practices. Int J Cancer 2002;98:262–7. [14] Whittemore AS, Kolonel LN, Wu AH, et al. Prostate cancer in relation to diet, physical activity, and body size in blacks, whites, and Asians in the United States and Canada. J Natl Cancer Inst 1995;87: 652– 61. [15] Talamini R, La Vecchia C, Decarli A, Negri E, Franceschi S. Nutrition, social factors and prostatic cancer in a northern Italian population. Br J Cancer 1986;53:817–21. [16] Ross RK, Shimizu H, Paganini-Hill A, Honda G, Henderson BE. Case-control studies of prostate cancer in blacks and whites in southern California. J Natl Cancer Inst 1987;78:869 –74. [17] Graham S, Haughey B, Marshall J, et al. Diet in the epidemiology of carcinoma of the prostate gland. J Natl Cancer Inst 1983;70: 687–92. [18] Kolonel LN, Yoshizawa CN, Hankin JH. Diet and prostatic cancer: a case-control study in Hawaii. Am J Epidemiol 1988;127:999 – 1012. [19] Rohan TE, Howe GR, Burch JD, Jain M. Dietary factors and risk of prostate cancer: a case-control study in Ontario, Canada. Cancer Causes Control 1995;6:145–54. [20] West DW, Slattery ML, Robison LM, French TK, Mahoney AW. Adult dietary intake and prostate cancer risk in Utah: a case-control study with special emphasis on aggressive tumors. Cancer Causes Control 1991;2:85–94. [21] Brawer MK, Bigler SA, Sohlberg OE, Nagle RB, Lange PH. Significance of prostatic intraepithelial neoplasia on prostate needle biopsy. Urology 1991;38:103–7. [22] Ohno Y, Yoshida O, Oishi K, Okada K, Yamabe H, Schroeder FH. Dietary b-carotene and cancer of the prostate: a case-control study in Kyoto, Japan. Cancer Res 1988;48:1331– 6. [23] Walker ARP, Walker BF, Tsotetsi NG, Sebitso C, Siwedi D, Walker AJ. Case-control study of prostate cancer in black patients in Soweto, South Africa. Br J Cancer 1992;65:438 – 41. [24] Talamini R, Franceschi S, La Vecchia C, Serraino D, Barra S, Negri E. Diet and prostate cancer: a case-control study in Northern Italy. Nutr Cancer 1992;18:277– 86. [25] Mettlin C, Selenskas S, Natarajan N, Huben R. Beta-carotene and animal fats and their relationship to prostate cancer risk. Cancer 1989;64:605–12. [26] Heshmat MY, Kaul L, Kovi J, et al. Nutrition and prostate cancer: a case-control study. Prostate 1985;6:7–17. [27] Mills PK, Beeson WL, Phillips RL, Fraser GE. Cohort study of diet, lifestyle, and prostate cancer in Adventist men. Cancer 1989;64:598 – 604. [28] Le Marchand L, Kolonel LN, Wilkens LR, Myers BC, Hirohata T. Animal fat consumption and prostate cancer: a prospective study in Hawaii. Epidemiology 1994;5:276 – 82. [29] Giovannucci E, Rimm EB, Colditz GA, et al. A prospective study of dietary fat and risk of prostate cancer. J Natl Cancer Inst 1993;85: 1571–9.
397
[30] Snowden DA, Phillips RL, Choi W. Diet, obesity, and risk of fatal prostate cancer. Am J Epidemiol 1984;120:244 –50. [31] Hsing AW, McLaughlin JK, Schuman LM, et al. Diet, tobacco use, and fatal prostate cancer: results from the Lutheran Brotherhood Cohort Study. Cancer Res 1990;50:6836 – 40. [32] Hirayama T. Epidemiology of prostate cancer with special reference to the role of diet. Natl Cancer Inst Monogr 1979;53:149 –55. [33] Terry PD, Rohan TE, Wolk A. Intakes of fish and marine fatty acids and the risks of cancers of breast and prostate and other hormone related cancers: a review of the epidemiological evidence. Am J Clin Nutr 2003;77:532– 43. [34] Augustsson K, Michaud DS, Rimm EB, et al. A prospective study of intake of fish and marine fatty acids and prostate cancer. Cancer Epidemiol Biomarkers Prev 2003;12:64 –7. [35] Terry P, Lichtenstein P, Feychting M, Ahlbom A, Wolk A. Fatty fish consumption and risk of prostate cancer. Lancet 2001;357:1764 – 6. [36] Harvei S, Bjerve KS, Tretli S, Jellum E, Robsahm TE, Vatten L. Prediagnostic level of fatty acids in serum phospholipids: omega-3 and omega-6 fatty acids and the risk of prostate cancer. Int J Cancer 1997;71:545–51. [37] Norrish AE, Skeaff CM, Arribas GL, Sharpe SJ, Jackson RT. Prostate cancer risk of consumption of fish oils: a dietary biomarker-based case-control study. Br J Cancer 1999;81:1238 – 42. [38] Astorg P. Dietary n-6 and n-3 polyunsaturated fatty acids and prostate cancer risk: a review of epidemiological and experimental evidence. Cancer Causes Control 2004;15:367– 86. [39] U.S. Environmental Protection Agency. Water Quality Conditions in the United States: A Profile from the 2000 National Water Quality Inventory. Washington, DC: U.S. EPA, 2001. [40] Morrison H, Savitz D, Semenciw R. Cancer among farmers. Occup Med 1991;6:335–54. [41] Committee to Review the Health Effects in Vietnam Veterans of Exposure to Herbicides (Division of Health Promotion and Disease Prevention, Institute of Medicine). Veterans and agent orange: update 1996. Washington, DC: National Academy Press, 1996. [42] Gann PH, Hennekens CH, Sacks FM, Grodstein F, Giovannucci EL, Stampfer MJ. Prospective study of plasma fatty acids and risk of prostate cancer. J Natl Cancer Inst 1994;86:281– 6. [43] Leitzmann MF, Stampfer MJ, Michaud DS, et al. Dietary intake of n-3 and n-6 fatty acids and risk of prostate cancer. Am J Clin Nutr 2004;80:204 –16. [44] West DW, Slattery ML, Robison LM, French TK, Mahoney AW. Adult dietary intake and prostate cancer risk in Utah: a case control with special emphasis on aggressive tumors. Cancer Causes Control 1991;2:85–94. [45] Ghadirian P, Lacroix A, Maisonneuve P, et al. Nutrition factors and prostate cancer risk: a case-control study of French Canadians in Montreal, Canada. Cancer Causes Control 1996;7:428 –36. [46] Brouwer IA, Katan MB, Zock PL. Dietary alpha-linolenic acid is associated with reduced risk of fatal coronary heart disease, but increased prostate cancer risk: a meta analysis. J Nutr 2004;134:919 – 22. [47] Kris-Etherton PM, Taylor DS, Yu-Poth S, et al. Polyunsaturated fatty acids in the food chain in the United States. Am J Clin Nutr 2000; 71(suppl.):179S– 88S. [48] Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macronutrients). Washington, DC: National Academy Press, 2000. [49] Grootveld M, Atherton MD, Sheerin AN, et al. In vivo adsorption, metabolism, and urinary excretion of alpha, beta-unsaturated aldehydes in experimental animals. Relevance to the development of cardiovascular diseases by the dietary ingestion of thermally stressed polyunsaturated-rich culinary oils. J Clin Invest 1998;101:1210 – 8. [50] Velasco J, Marmesat S, Bordeaux O, Marquez-Ruiz G, Dobarganes C. Formation and evolution of monoepoxy fatty acids in thermoxidized olive and sunflower oils and quantitation in used frying oils
398
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
from restaurants and fried-food outlets. J Agric Food Chem 2004; 52:4438 – 43. [51] Indart A, Viana M, Grootveld MC, Silwood CJ, Sanchez-Vera I, Bonet B. Tetratogenic actions of thermally-stressed culinary oils in rats. Free Radical Res 2002;36:1051– 8. [52] Yang CM, Kendall CW, Stamp D, Medline A, Archer MC, Bruce WR. Thermally oxidized dietary fat and colon carcinogenesis in rodents. Nutr Cancer 1998;30:69 –73. [53] Chao PM, Chao CY, Lin FJ, Huang C. Oxidized frying oil up-
regulates acyl-CoA oxidase P450 4 A1 genes in rats and activates PPARalpha. J Nutr 2001;131:3166 –74. [54] Sulzie A, Hirche F, Eder K. Thermally oxidized dietary fat upregulates the expression of target genes of PPAR alpha in rat liver. J Nutr 2004;134:1375– 83. [55] Sawa T, Akaike T, Kida K, Fukushima Y, Takagi K, Maeda H. Lipid peroxyl radicals from oxidized oils and heme-iron: implications of a high fat diet in carcinogenesis. Cancer Epidemiol Biomarkers Prev 1998;7:1007–12.
Original article
Comparisons of prostate cancer mortality rates with dietary practices in the United States Janet Laura Colli, M.D.*, Albert Colli, B.S.Ch.E. Division of Urology, Department of Surgery, University of Alabama at Birmingham, Birmingham, AL, USA Received 25 November 2004; received in revised form 21 February 2005; accepted 3 March 2005
Abstract From 1930 to 1992, prostate cancer mortality rates in the United States doubled and then declined somewhat until 2000. The objective of this study is to determine whether variations in prostate cancer mortality rates correlate with dietary changes that occurred over that period. Simple linear regression models were applied to age-adjusted prostate cancer mortality rates and per-capita consumption rates for 18 foods from 1930 to 2000. Correlation coefficients were calculated while comparing food consumption rates to prostate cancer mortality rates for the same year. Correlation coefficients were then recalculated when the prostate cancer mortality rates were compared with food consumption rates that occurred: 1 yr; 2 yr; 3 yr; and continuing in progression for 21 yr before the occurrence of the prostate cancer mortality. The largest positive correlation coefficients were associated with the consumption of: total meat (red meat, poultry and fish) (R ⫽ 0.83, T between 0 and 1); added fats and oils (R ⫽ 0.83, T ⫽ 21); ice cream (R ⫽ 0.83, T ⫽ 20); margarine (R ⫽ 0.81, T ⫽ 4); salad/cooking oil (R ⫽ 0.82, T between 3 and 4) and; vegetable shortening (R ⫽ 0.81, T between 1 and 2) where R is the correlation coefficient and T is the time in years between consumption and mortality. In conclusion, this study found strong positive correlations between prostate cancer mortality and the consumption of: total meat; added fats and oils, ice cream, salad/cooking oils, margarine, and vegetable shortening. The connection between total meat consumption and prostate cancer risk is consistent with previous studies in the literature. The link between salad/cooking oil consumption and prostate cancer risk may be consistent with past studies which suggest that -linolenic acid (a component of salad/cooking oils) is a suspected risk factor for prostate cancer. © 2005 Elsevier Inc. All rights reserved. Keywords: Prostate cancer; Foods; Diet; Epidemiology
1. Introduction In the period from 1930 to 2000 prostate cancer mortality rates in the United States varied considerably. Over that period, there have been changes in dietary practices as people acquired different tastes and consumed foods that became commercially available. The purpose of this study is to compare age-adjusted prostate cancer mortality to the per-capita consumption of various foods to determine if there are correlations that suggest potential associations. Fortunately, average per-capita consumption levels for various foods, based on food disappearance data, have been compiled by the United States Department of Agriculture since 1909 [1]. More recently, a study from an American Cancer Society publication presented age-adjusted cancer
* Corresponding author. Tel.: ⫹1-214-707-8801; fax: ⫹1-205-9341470. E-mail address: [email protected] (J.L. Colli). 1078-1439/05/$ – see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.urolonc.2005.03.020
mortality for cancers at various sites (including prostate cancer) for the period from 1930 to 2000 [2]. The existence of these two data bases allowed this study to be conducted. In this study, we selected prostate cancer mortality rather than incidence as the focus of the analysis for several reasons. Incidence data are only available for a much more limited time period [3] than mortality data reducing the robustness of the analysis. In addition, prostate-specific antigen testing introduced in the late 1980s increased the number of prostate cancer incidences detected and would have been a confounding factor to the dietary factors analyzed in this study. Although there has been a decline in prostate cancer mortality rates since the early 1990s, this decrease does not appear to be associated with early detection from prostate-specific antigen testing and treatment [4]. There are other reasons other than dietary changes that may have caused changes in prostate cancer mortality over that 71 yr period: toxic contaminants have been introduced into the environment because of industrialization; ethnic
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398 Table 1 Cross correlation coefficients between prostate cancer mortality rates and per-capita consumption rates of animal products in the United States from 1930 to 2000 Time* (years)
Total meat
Red meat
Beef
Pork
Chicken
Fish
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
0.86 0.86 0.85 0.85 0.84 0.82 0.81 0.79 0.79 0.79 0.78 0.78 0.77 0.77 0.77 0.76 0.76 0.74 0.73 0.71 0.70 0.69
0.67 0.69 0.7 0.71 0.72 0.72 0.72 0.71 0.72 0.73 0.74 0.74 0.74 0.75 0.75 0.75 0.75 0.74 0.72 0.70 0.69 0.67
0.68 0.70 0.72 0.73 0.74 0.74 0.75 0.75 0.75 0.76 0.76 0.76 0.75 0.75 0.74 0.74 0.74 0.73 0.72 0.70 0.68 0.64
0.5 0.48 0.45 0.43 0.44 0.43 0.38 0.34 0.35 0.37 0.39 0.39 0.40 0.43 0.41 0.41 0.41 0.42 0.41 0.39 0.42 0.45
0.79 0.78 0.78 0.77 0.76 0.75 0.73 0.73 0.73 0.73 0.74 0.74 0.73 0.73 0.72 0.70 0.69 0.67 0.65 0.65 0.64 0.64
0.78 0.75 0.71 0.67 0.63 0.60 0.57 0.55 0.50 0.48 0.42 0.37 0.35 0.32 0.30 0.28 0.25 0.24 0.22 0.20 0.16 0.10
* Time in years from consumption to mortality.
changes in the population have occurred from migration; pesticide use has increased and; hormones have been introduced into the food supply to increase the growth of animals for human consumption. Nevertheless, examining the relationship between prostate cancer mortality and dietary practices over such a long period of time as 71 yr may provide some insight into the causes of prostate cancer.
2. Methods 2.1. Data sources Age-standardized prostate cancer mortality rates from 1930 to 2000 were obtained from Jemal et al. [2] published by the American Cancer Society. The authors of the study obtained mortality data from 1930 to 2001 from the National Center for Health Statistics [5]. In their study, causes of death were coded and classified according to the International Classification of Diseases (ICD-8, ICD-9, and ICD10) [6 – 8]. The paper presented age-adjusted mortality rates that were standardized to the 2000 United States standard population and expressed per 100,000 population [2]. Per-capita food consumption rates from 1920 to 2000 were obtained from the food consumption data system [1] compiled by the Economic Research Service (ERS) for the United States Department of Agriculture. Information on domestic production, imports, exports and losses were used
391
by ERS statisticians in estimating average per-capita consumption levels for foods in the database. ERSs food availability data represent the food supply, or the disappearance of food into the food marketing system, because they are normally calculated as the residual of a commodity’s total annual available supply after subtracting measurable uses, such as farm inputs (feed and seed), exports, ending stocks, and industrial uses. For this study we selected per-capita consumption rates for the following foods, or food categories for which data was available for the period from 1920 to 2000: total meat (red meat, poultry, fish), red meat, beef, pork, chicken, fish, dairy, milk, cheese, ice cream, eggs, flour and cereal, fats and oils, butter, margarine, salad/ cooking oil, shortening (animal fat and vegetable shortening) and vegetable shortening. In all cases except eggs, the units of the food consumption rates are provided in kilograms per year. The units for eggs are: whole eggs per year. 2.2. Statistical analysis Simple linear regression models were applied to prostate cancer mortality rates and per-capita consumption rates for eighteen foods (listed above) for the period from 1930 to 2000. Correlation coefficients were calculated when food consumption rates were compared to prostate cancer mortality rates for the same year. Correlation coefficients were then recalculated when the prostate cancer mortality rates were compared with food consumption rates that occurred:
Table 2 Cross correlation coefficients between prostate cancer mortality rates and per-capita consumption rates of dairy products, eggs and flour in the United States from 1930 to 2000 Time* (years)
Dairy
Milk
Cheese
Ice cream
Eggs
Flour
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
⫺0.49 ⫺0.46 ⫺0.43 ⫺0.41 ⫺0.40 ⫺0.41 ⫺0.40 ⫺0.39 ⫺0.37 ⫺0.33 ⫺0.30 ⫺0.24 ⫺0.22 ⫺0.16 ⫺0.09 ⫺0.02 0.03 0.06 0.09 0.14 0.18 0.20
⫺0.59 ⫺0.56 ⫺0.54 ⫺0.53 ⫺0.52 ⫺0.53 ⫺0.53 ⫺0.52 ⫺0.51 ⫺0.48 ⫺0.47 ⫺0.42 ⫺0.41 ⫺0.36 ⫺0.29 ⫺0.22 ⫺0.18 ⫺0.16 ⫺0.16 ⫺0.11 ⫺0.08 ⫺0.07
0.79 0.79 0.79 0.78 0.78 0.77 0.76 0.76 0.75 0.74 0.73 0.72 0.71 0.71 0.71 0.71 0.71 0.70 0.71 0.71 0.70 0.69
0.67 0.68 0.68 0.67 0.67 0.67 0.68 0.68 0.69 0.70 0.71 0.72 0.72 0.73 0.74 0.76 0.78 0.80 0.81 0.83 0.83 0.83
⫺0.50 ⫺0.50 ⫺0.50 ⫺0.50 ⫺0.51 ⫺0.50 ⫺0.50 ⫺0.48 ⫺0.45 ⫺0.40 ⫺0.36 ⫺0.33 ⫺0.28 ⫺0.23 ⫺0.20 ⫺0.16 ⫺0.11 ⫺0.06 ⫺0.02 ⫺0.01 0.03 0.09
⫺0.45 ⫺0.51 ⫺0.56 ⫺0.61 ⫺0.66 ⫺0.70 ⫺0.74 ⫺0.78 ⫺0.80 ⫺0.81 ⫺0.82 ⫺0.84 ⫺0.85 ⫺0.87 ⫺0.88 ⫺0.89 ⫺0.89 ⫺0.90 ⫺0.90 ⫺0.91 ⫺0.91 ⫺0.91
* Time in years from consumption to mortality.
392
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
Table 3 Cross correlation coefficients between prostate cancer mortality rates and per-capita consumption rates for added fats in the United States from 1930 to 2000 Time* (years)
Fats/oils
Butter
Margarine
Salad and cooking oil
Shortening
Vegetable shortening
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
0.72 0.73 0.72 0.71 0.71 0.70 0.70 0.70 0.72 0.74 0.75 0.74 0.73 0.73 0.73 0.74 0.75 0.77 0.79 0.81 0.82 0.83
⫺0.87 ⫺0.87 ⫺0.86 ⫺0.86 ⫺0.85 ⫺0.84 ⫺0.83 ⫺0.82 ⫺0.82 ⫺0.81 ⫺0.80 ⫺0.78 ⫺0.76 ⫺0.75 ⫺0.75 ⫺0.75 ⫺0.75 ⫺0.75 ⫺0.74 ⫺0.75 ⫺0.76 ⫺0.76
0.80 0.80 0.80 0.81 0.81 0.81 0.80 0.80 0.80 0.81 0.80 0.79 0.79 0.78 0.78 0.79 0.79 0.79 0.80 0.80 0.80 0.80
0.81 0.82 0.82 0.82 0.82 0.81 0.81 0.81 0.80 0.80 0.81 0.81 0.80 0.81 0.81 0.81 0.80 0.80 0.81 0.81 0.80 0.79
0.34 0.31 0.28 0.24 0.22 0.19 0.17 0.16 0.17 0.21 0.20 0.11 0.02 ⫺0.03 ⫺0.07 ⫺0.12 ⫺0.16 ⫺0.15 ⫺0.09 ⫺0.14 0.25 0.36
0.80 0.81 0.81 0.80 0.78 0.77 0.77 0.77 0.79 0.79 0.79 0.76 0.74 0.71 0.69 0.67 0.67 0.68 0.69 0.70 0.70 0.70
* Time in years from consumption to mortality.
1 yr; 2 yr; 3 yr; and continuing in progression for 21 yr before occurrence of the prostate cancer mortality. Cross correlation coefficients of the independent variables were calculated to identify collinearity (linear or nearly linear relationships) between the consumption variables that exhibited a strong correlation with prostate cancer mortality. Collinearity between the independent variables causes problems for step-wise multiple regression analysis and makes it difficult to make any conclusions from the results. All the data in this study was analyzed using the software program SPSS® Base for Windows®, version 12.0, by SPSS, Inc. (Chicago, IL).
3. Results Correlation coefficients from regressing prostate cancer mortality against independent variables (foods, or food categories) are shown in Tables 1 through 3. The largest positive correlation coefficients were associated with the consumption of: total meat (red meat, poultry and fish) from zero to 1 yr before the prostate cancer mortality (R ⫽ 0.86); added fats and oils 21 yr before mortality (R ⫽ 0.83); ice cream 21 yr before mortality (R ⫽ 0.830); margarine 4 yr before mortality (R ⫽ 0.81); salad and cooking oil from 2 to 3 yr before mortality (R ⫽ 0.82); and vegetable shortening from 1 to 2 yr before mortality (R ⫽ 0.81) where R is the correlation coefficient. The largest negative correlation coefficients were associated with the consumption of flour 19
yr before mortality (R ⫽ ⫺0.91); and butter from zero to 1 yr before mortality (R ⫽ ⫺0.87). Neither red meat (R ⫽ 0.67), poultry (R ⫽ 0.80) or fish (R ⫽ 0.78) individually had a correlation coefficient that was as high as the total meat category. U.S. per-capita total meat consumption rates and age-adjusted prostate cancer mortality rates for the period 1930 to 2000 are shown in Fig. 1. Total meat consumption rates versus age-adjusted prostate cancer mortality rates for the period 1930 to 2000 are shown in Fig. 2. On this graph, each point corresponds to an annual prostate cancer mortality rate mated with the total meat consumption rate that occurred the previous year. Salad/cooking oil and vegetable oil had the next highest positive correlation coefficient of the independent variables investigated. U.S. per-capita salad/cooking oil and vegetable oil consumption rates together with age-adjusted prostate cancer mortality rates for the period 1930 to 2000 are shown in Fig. 3. Salad/cooking oil consumption rates versus age-adjusted prostate cancer mortality rates for the period 1930 to 2000 are shown in Fig. 4. On this graph, each point corresponds to an annual prostate cancer mortality rate mated with a salad/cooking oil consumption rate that occurred 3 yr prior. Vegetable shortening and margarine both had high positive correlation coefficients with prostate cancer mortality whereas butter had a strong negative correlation coefficient. U.S. per-capita vegetable shortening, margarine and butter consumption rates together with age-adjusted prostate can-
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
393
between the consumption variables. The correlation matrix from this analysis generated the following correlation coefficients between the independent variables: total meat, salad/ cooking oil (R ⫽ 0.95); total meat, butter (R ⫽ ⫺0.94); margarine, butter (R ⫽ ⫺0.96); salad/cooking oil, shortening (R ⫽ 0.96); salad/cooking oil, vegetable shortening (R ⫽ 0.97). These results indicate that some of the independent variables that correlated closely with prostate cancer mortality were collinear and highly interconnected. Collinearity between the independent variables makes it difficult to make any conclusions from the results of stepwise multiple regression analysis. As a result, a step-wise multiple regression analysis was not performed to develop a mathematical model for the relationship between the independent variables and prostate cancer mortality.
4. Discussion
Fig. 1. U.S. prostate cancer mortality and per-capita annual total meat consumption (Kg/y).
cer mortality rates for the period 1930 to 2000 are shown in Fig. 5. Cross correlation coefficients of the independent variables were calculated to determine the degree of collinearity
In this study, the strongest dietary factors that correlated with increased prostate cancer mortality were the consumption of: total meat (that includes red meat, poultry, and fish); and certain vegetable oils categories (total added fats and oils, margarine, salad/cooking oils, and vegetable shortening). The category “total meat” results in the consumption of animal fats that are contained in red meat, poultry and fish. Ecological correlation studies in the 1970s showed a positive association between saturated fat consumption and prostate cancer incidence or mortality [9 –11]. Armstrong and Doll [9] hypothesized that dietary fat may be a major cause of prostate cancer based on a correlation coefficient of 0.74 between national per-capita consumption fat and pros-
Fig. 2. U.S. per-capita total meat consumption versus prostate cancer mortality (1930 –2000).
394
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
Fig. 3. U.S. prostate cancer mortality and per-capita salad/cooking oil consumption (Kg/y).
tate cancer mortality. More recently, an international study [12] comparing per-capita food consumption rates and prostate cancer mortality rates from different countries reported correlation coefficients of 0.78 for animal fat. In a similar international study [13], the correlation coefficient between the consumption of animal fats and prostate cancer risk were found to be 0.61. Many case-control studies have examined the association between animal fat and prostate cancer [14 –26]. Although the 13 studies differed in terms of study design, only four studies [19,22,25,26] failed to show a positive association with total fat intake. The association between fat intake and prostate cancer has been investigated in six cohort studies [27–32]. A positive association was reported in four of the studies [27–30] while two other studies did not detect an association [31,32]. Dietary fat intake from animal sources has been more consistently linked to prostate cancer than any other food category. These past studies, taken in totality, support an association between animal fats and prostate cancer risk. The high correlation coefficient between prostate cancer risk and fish consumption found in the current study runs counter to some of the evidence in the literature. Although the consumption of saturated fats have been associated with
increased prostate cancer risk, experimental studies suggest that marine fatty acids, particularly the long-chained eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) inhibit the proliferation of prostate cancer cells in vitro and reduce the progression of these tumors in laboratory animals [33]. Whether the consumption of marine fatty acids from fish can reduce prostate cancer risk in humans is unclear. In a recent prospective study of 47,882 men, eating fish more than three times per week was associated with a reduced risk of prostate cancer, and the strongest association was with metastatic cancer [34]. In a population-based cohort study of 6,272 Swedish men with 30 yr of follow-up, men who ate no fish had a two to three fold increase of prostate cancer than those who ate moderate or high amounts [35]. However, a cohort study of prostate cancer in Adventist men showed several trends of increasing risk of prostate cancer mortality with the consumption of fish [27]. In a case-control study of blood donors to a serum data bank who later developed prostate cancer, and a matched set without prostate cancer, there was no clear association between risk of prostate cancer and the presence of EPA and DHA fatty acids in the blood [36]. In another case-control study, reduced prostate cancer risk was associated with high levels of erythrocyte biomarkers for EPA and DHA [37]. In two international studies, fish consumption rates from many countries were compared to prostate cancer mortality rates to determine whether there is a correlation. In one study, the correlation coefficient for the period from 1961 to 1990 was 0.015 indicating almost no association [13]. In the other study, the correlation coefficient was ⫺0.39 suggesting that fish consumption may reduce prostate cancer risk [12]. A recent review of the literature concluded that current epidemiological evidence provides little support that marine fatty acids provide protection from prostate cancer risk [38]. Although existing epidemiological studies regarding the consumption of marine fatty acids and prostate cancer risk are inconclusive, there is little evidence that consumption of fish is associated with increased prostate cancer risk as suggested by the current study. However, reports of toxic contamination of domestic fish and shellfish are a potential cause for an association between fish consumption and increased prostate cancer risk. From 1992 to 1998, fish samples were collected from 233 stream sites by the U.S. Geological Survey’s National Water Quality program. Tissue composites from whole fish were analyzed for mercury, polychlorinated biphenols (PCBs), chlorodane, dioxane (2,3,7,8-tetrachlorodibenzo-p-dioxin) and dichlorodiphenyltrichlorethane (DDT), and trace elements. More than 90% of the sampled fish had at least one contaminant and about half had at least five contaminants [39]. The U.S. Environmental Protection Agency’s National Listing of Fish and Wildlife Advisories listed 2,838 advisories that were issued by state and local officials to protect public health in 2000; mercury, PCBs, chlordane, dioxane and DDT were responsible for 99% of the consumption advisories [39]. The Department of Health and Human Services
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
395
Fig. 4. U.S. per-capita salad and cooking oil consumption versus prostate cancer mortality (1930 –2000).
has concluded that PCBs, dioxane and DDT may reasonably be anticipated to be carcinogens. One study has implicated organochlorine pesticides, similar to those found as contaminants in fish and shellfish, as the cause for the increased rate of prostate cancer among farm workers [40]. The National Academy of Science’s committee to review the effects of exposure to herbicides in Vietnam veterans concluded that there was limited suggestive evidence linking herbicide exposure to prostate cancer [41]. In this current study, the consumption of salad/cooking oils correlated with increased prostate cancer mortality. Salad/cooking oils are a major source of -linolenic acid (an essential polyunsaturated fatty acid) in the U.S. diet. Giovannucci et al. [29] and Gann et al. [42] found: a strong positive association between -linolenic acid and prostate cancer risk; no clear linear relation across quartiles of exposure suggesting a threshold effect; that low levels of linoleic acid, and another polyunsaturated fatty acid, may further exaggerate the effect. More recently, in a cohort study of 47,866 men, Leitzmann et al. [43] found that -linolenic acid was unrelated to total prostate cancer risk but strongly related to advanced prostate cancer risk. Several case-controls studies [44,45] suggest an association with the consumption of -linolenic acid and prostate cancer risk. The author of a recent review of the literature [38] and a meta-analysis [46] concluded that -linolenic acid may be a risk factor for prostate cancer. From 1930 to 2000, salad/cooking oil consumption in the U.S. increased from 6.2 to 35.2 pounds per year resulting in an increase in per-capita -linolenic acid intake. It has been estimated that per-capita -linolenic acid consumption in the U.S. was 1.4 grams per day in 1990 [47]. This is based primarily on the consumption of soybean and canola oil,
which contain high levels of -linolenic acid: 7.4% and 9.8%, respectively [47]. Per-capita -linolenic acid consumption from vegetable oils is estimated to be about 0.1 grams per day in 1930, or about 10% of the level at the end of the 20th century. The margarine and vegetable shortening categories, both of which contain low levels of -linolenic acid [47] but high levels of hydrogenated polyunsaturated fatty acids or trans fatty acids [48], correlate strongly with increasing prostate cancer mortality. No information was found in the literature on the relationship between trans fatty acid consumption and prostate cancer risk although the National Academy of Science has recommended minimizing the consumption of trans fatty acids since they provide no benefit to human health [48]. Shortening, but not margarine, is often thermally stressed because it is used to fry foods at elevated temperatures. Thermally stressing polyunsaturated fatty acids found in vegetable shortening during routine frying or cooking, generates high levels of oxidized dietary fat containing both cytotoxic aldehydic products [49] and monoepoxy fatty acids [50]. It should be noted that there was a high correlation coefficient between prostate cancer mortality and vegetable shortening but not shortening that also contain saturated animal fats. Oxidized dietary fats have been reported to exert teratogenic actions on experimental animals [51], colon carcinogenicity in rodents [52], suppress gene expression of lipogenic enzymes in the liver of rats [53,54] and generate components that exert DNA-cleaving activity [55]. No information was found in the literature regarding the relationship between thermally stressed polyunsaturated fatty acids and prostate cancer risk. Butter was found to correlate negatively with prostate
396
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
Fig. 5. U.S. prostate cancer mortality rate and vegetable shortening, margarine or butter consumption.
cancer risk. However, it is not clear that this correlation results from protections from prostate cancer risk provided by butter consumption or by decreased risk from reduced margarine consumption. Butter and margarine are strongly negatively interrelated; the correlation coefficient between the two independent variables is ⫺0.96. The strong negative correlation between prostate cancer mortality and flour consumption found in this study is consistent with the results of two international studies [12,13] comparing per-capita food consumption rates and prostate cancer mortality rates from different countries. The maximum correlation coefficient occurred when the time period between consumption and prostate cancer mortality was: from zero to 1 yr for total meat; 21 yr for added fats and oils; 20 yr for ice cream; 4 yr for margarine; from 2 to 3 yr for salad/cooking oil; and from 1 to 2 yr for vegetable shortening. This would suggest that if the association was real and not coincidental, these time periods would represent the latency period between exposure and prostate cancer mortality. There is a high degree of uncertainty in these time estimates since it is not clear that small
changes in the value of the regression coefficient (which established the peak correlation in many cases) are significant. The correlation coefficients between prostate cancer mortality and several food categories identified in this study are very high: 0.94 between total meat and salad/cooking oils; and 0.97 between vegetable shortening and salad/cooking oils. This means that many of the independent variables that correlate strongly with prostate cancer mortality are highly interdependent suggesting that only some of them may actually affect prostate cancer risk. This study has many sources of uncertainty. For example, per-capita food intake levels are obtained from food that is produced and imported, reduced by food that is exported and wasted. It does not account completely for food wastage, or foods fed to pets although animal feed and seed usage on farms is taken into consideration. Also, food disappearance levels represent averages for the population and do not take into consideration variations that can occur throughout the country because of affluence or ethnicity. More importantly, differences in food consumption patterns between men having advanced prostate cancer and the averages for the population (as used in this study) are not known. Since the results of this study have a high degree of uncertainty, they should not be used for dietary recommendations but may provide additional insights into risk factors for prostate cancer that need to be corroborated or disproved by further research. In particular, the results from this study suggest further research into the relationship between prostate cancer risk and the consumption of: -linolenic acid; trans fatty acids and; thermally stressed polyunsaturated fatty acids.
References [1] Economic Research Service, United States Department of Agriculture. Food consumption (per capita) data system. Available at: http:// www.ers.usda.gov./Data/FoodConsumption/. Accessed July 2004. [2] Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. Cancer J Clin 2004;54:8 –29. [3] Surveillance, Epidemiology, and End Results (SEER) Program (http://www.seer.cancer.gov/) SEER*Stat Database: Incidence SEER 9 Regs Public-Use, Nov 2002 Sub (1973–2000), National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch. [4] Brenner H, Arndt V. Long-term survival rates of patients with prostate cancer-specific antigen screening era: population-based estimates for the year 2000 by period analysis. J Clin Oncol 2005;23:441–7. [5] National Center for Health Statistics, Division of Vital Statistics, Centers for Disease Control. Available at: http://www.cdc.gov/nchs/ nvss.htm. Accessed July 2004. [6] Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death. Vol. 1, 10th revision. Geneva: World Health Organization, 1992. [7] Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death. Vol. 1, 9th revision. Geneva: World Health Organization, 1975.
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398 [8] Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death. Vol. 1, 8th revision. Geneva: World Health Organization, 1967. [9] Armstrong B, Doll R. Environmental factors and cancer incidence and mortality in different countries, with special references to dietary practices. Int J Cancer 1975;15:617–31. [10] Howell MA. Factor analysis of international cancer mortality data and per capita food consumption. Br J Cancer 1974;29:328 –36. [11] Blair A, Fraumeni JF Jr. Geographic patterns of prostate cancer in the United States. J Natl Cancer Inst 1978;61:379 – 84. [12] Hebert JR, Hurley TG, Olendski BC, Teas J, Ma Y, Hampl JS. Nutritional and Socioeconomic factors in relation to prostate cancer mortality: a cross national study. J Natl Cancer Inst 1998;90:1637– 47. [13] Ganmaa D, Li XM, Wang J, Qin LQ, Wang PY, Sato A. Incidence and mortality of testicular and prostatic cancers in relation to world dietary practices. Int J Cancer 2002;98:262–7. [14] Whittemore AS, Kolonel LN, Wu AH, et al. Prostate cancer in relation to diet, physical activity, and body size in blacks, whites, and Asians in the United States and Canada. J Natl Cancer Inst 1995;87: 652– 61. [15] Talamini R, La Vecchia C, Decarli A, Negri E, Franceschi S. Nutrition, social factors and prostatic cancer in a northern Italian population. Br J Cancer 1986;53:817–21. [16] Ross RK, Shimizu H, Paganini-Hill A, Honda G, Henderson BE. Case-control studies of prostate cancer in blacks and whites in southern California. J Natl Cancer Inst 1987;78:869 –74. [17] Graham S, Haughey B, Marshall J, et al. Diet in the epidemiology of carcinoma of the prostate gland. J Natl Cancer Inst 1983;70: 687–92. [18] Kolonel LN, Yoshizawa CN, Hankin JH. Diet and prostatic cancer: a case-control study in Hawaii. Am J Epidemiol 1988;127:999 – 1012. [19] Rohan TE, Howe GR, Burch JD, Jain M. Dietary factors and risk of prostate cancer: a case-control study in Ontario, Canada. Cancer Causes Control 1995;6:145–54. [20] West DW, Slattery ML, Robison LM, French TK, Mahoney AW. Adult dietary intake and prostate cancer risk in Utah: a case-control study with special emphasis on aggressive tumors. Cancer Causes Control 1991;2:85–94. [21] Brawer MK, Bigler SA, Sohlberg OE, Nagle RB, Lange PH. Significance of prostatic intraepithelial neoplasia on prostate needle biopsy. Urology 1991;38:103–7. [22] Ohno Y, Yoshida O, Oishi K, Okada K, Yamabe H, Schroeder FH. Dietary b-carotene and cancer of the prostate: a case-control study in Kyoto, Japan. Cancer Res 1988;48:1331– 6. [23] Walker ARP, Walker BF, Tsotetsi NG, Sebitso C, Siwedi D, Walker AJ. Case-control study of prostate cancer in black patients in Soweto, South Africa. Br J Cancer 1992;65:438 – 41. [24] Talamini R, Franceschi S, La Vecchia C, Serraino D, Barra S, Negri E. Diet and prostate cancer: a case-control study in Northern Italy. Nutr Cancer 1992;18:277– 86. [25] Mettlin C, Selenskas S, Natarajan N, Huben R. Beta-carotene and animal fats and their relationship to prostate cancer risk. Cancer 1989;64:605–12. [26] Heshmat MY, Kaul L, Kovi J, et al. Nutrition and prostate cancer: a case-control study. Prostate 1985;6:7–17. [27] Mills PK, Beeson WL, Phillips RL, Fraser GE. Cohort study of diet, lifestyle, and prostate cancer in Adventist men. Cancer 1989;64:598 – 604. [28] Le Marchand L, Kolonel LN, Wilkens LR, Myers BC, Hirohata T. Animal fat consumption and prostate cancer: a prospective study in Hawaii. Epidemiology 1994;5:276 – 82. [29] Giovannucci E, Rimm EB, Colditz GA, et al. A prospective study of dietary fat and risk of prostate cancer. J Natl Cancer Inst 1993;85: 1571–9.
397
[30] Snowden DA, Phillips RL, Choi W. Diet, obesity, and risk of fatal prostate cancer. Am J Epidemiol 1984;120:244 –50. [31] Hsing AW, McLaughlin JK, Schuman LM, et al. Diet, tobacco use, and fatal prostate cancer: results from the Lutheran Brotherhood Cohort Study. Cancer Res 1990;50:6836 – 40. [32] Hirayama T. Epidemiology of prostate cancer with special reference to the role of diet. Natl Cancer Inst Monogr 1979;53:149 –55. [33] Terry PD, Rohan TE, Wolk A. Intakes of fish and marine fatty acids and the risks of cancers of breast and prostate and other hormone related cancers: a review of the epidemiological evidence. Am J Clin Nutr 2003;77:532– 43. [34] Augustsson K, Michaud DS, Rimm EB, et al. A prospective study of intake of fish and marine fatty acids and prostate cancer. Cancer Epidemiol Biomarkers Prev 2003;12:64 –7. [35] Terry P, Lichtenstein P, Feychting M, Ahlbom A, Wolk A. Fatty fish consumption and risk of prostate cancer. Lancet 2001;357:1764 – 6. [36] Harvei S, Bjerve KS, Tretli S, Jellum E, Robsahm TE, Vatten L. Prediagnostic level of fatty acids in serum phospholipids: omega-3 and omega-6 fatty acids and the risk of prostate cancer. Int J Cancer 1997;71:545–51. [37] Norrish AE, Skeaff CM, Arribas GL, Sharpe SJ, Jackson RT. Prostate cancer risk of consumption of fish oils: a dietary biomarker-based case-control study. Br J Cancer 1999;81:1238 – 42. [38] Astorg P. Dietary n-6 and n-3 polyunsaturated fatty acids and prostate cancer risk: a review of epidemiological and experimental evidence. Cancer Causes Control 2004;15:367– 86. [39] U.S. Environmental Protection Agency. Water Quality Conditions in the United States: A Profile from the 2000 National Water Quality Inventory. Washington, DC: U.S. EPA, 2001. [40] Morrison H, Savitz D, Semenciw R. Cancer among farmers. Occup Med 1991;6:335–54. [41] Committee to Review the Health Effects in Vietnam Veterans of Exposure to Herbicides (Division of Health Promotion and Disease Prevention, Institute of Medicine). Veterans and agent orange: update 1996. Washington, DC: National Academy Press, 1996. [42] Gann PH, Hennekens CH, Sacks FM, Grodstein F, Giovannucci EL, Stampfer MJ. Prospective study of plasma fatty acids and risk of prostate cancer. J Natl Cancer Inst 1994;86:281– 6. [43] Leitzmann MF, Stampfer MJ, Michaud DS, et al. Dietary intake of n-3 and n-6 fatty acids and risk of prostate cancer. Am J Clin Nutr 2004;80:204 –16. [44] West DW, Slattery ML, Robison LM, French TK, Mahoney AW. Adult dietary intake and prostate cancer risk in Utah: a case control with special emphasis on aggressive tumors. Cancer Causes Control 1991;2:85–94. [45] Ghadirian P, Lacroix A, Maisonneuve P, et al. Nutrition factors and prostate cancer risk: a case-control study of French Canadians in Montreal, Canada. Cancer Causes Control 1996;7:428 –36. [46] Brouwer IA, Katan MB, Zock PL. Dietary alpha-linolenic acid is associated with reduced risk of fatal coronary heart disease, but increased prostate cancer risk: a meta analysis. J Nutr 2004;134:919 – 22. [47] Kris-Etherton PM, Taylor DS, Yu-Poth S, et al. Polyunsaturated fatty acids in the food chain in the United States. Am J Clin Nutr 2000; 71(suppl.):179S– 88S. [48] Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macronutrients). Washington, DC: National Academy Press, 2000. [49] Grootveld M, Atherton MD, Sheerin AN, et al. In vivo adsorption, metabolism, and urinary excretion of alpha, beta-unsaturated aldehydes in experimental animals. Relevance to the development of cardiovascular diseases by the dietary ingestion of thermally stressed polyunsaturated-rich culinary oils. J Clin Invest 1998;101:1210 – 8. [50] Velasco J, Marmesat S, Bordeaux O, Marquez-Ruiz G, Dobarganes C. Formation and evolution of monoepoxy fatty acids in thermoxidized olive and sunflower oils and quantitation in used frying oils
398
J.L. Colli, A. Colli / Urologic Oncology: Seminars and Original Investigations 23 (2005) 390 –398
from restaurants and fried-food outlets. J Agric Food Chem 2004; 52:4438 – 43. [51] Indart A, Viana M, Grootveld MC, Silwood CJ, Sanchez-Vera I, Bonet B. Tetratogenic actions of thermally-stressed culinary oils in rats. Free Radical Res 2002;36:1051– 8. [52] Yang CM, Kendall CW, Stamp D, Medline A, Archer MC, Bruce WR. Thermally oxidized dietary fat and colon carcinogenesis in rodents. Nutr Cancer 1998;30:69 –73. [53] Chao PM, Chao CY, Lin FJ, Huang C. Oxidized frying oil up-
regulates acyl-CoA oxidase P450 4 A1 genes in rats and activates PPARalpha. J Nutr 2001;131:3166 –74. [54] Sulzie A, Hirche F, Eder K. Thermally oxidized dietary fat upregulates the expression of target genes of PPAR alpha in rat liver. J Nutr 2004;134:1375– 83. [55] Sawa T, Akaike T, Kida K, Fukushima Y, Takagi K, Maeda H. Lipid peroxyl radicals from oxidized oils and heme-iron: implications of a high fat diet in carcinogenesis. Cancer Epidemiol Biomarkers Prev 1998;7:1007–12.