Index-Based Dietary Patterns and the Risk of Prostate Cancer among Iranian Men

Background and objective: The second most common cancer in men after lung cancer is prostate cancer (PC). Previous studies assessed the association between food items or food groups and the risk of PC, but diet quality indices are unique approaches to study any relations between diet and disease. Our objective was to investigate the effect of healthy eating index (HEI-2010) and Mediterranean-Style Dietary Pattern Score (MSDPS) on PC risk. Methods: In this case-control study, we recruited 97 patients with MS and 205 control subjects . Dietary intake was evaluted using a valid and reliable food frequency questionnaire. The HEI and MSDPS were calculated. Logistic regression was used to evaluate the relationship between HEI and MSDP scores and PC risk after adjusting the confounders. Results: In comparison to controls, cases had lower score on HEI (61 vs. 70.07; P< 0.001), and higher score on MSDP (26.20 vs. 24.49; P= 0.44). After comparing the highest and the lowest tertile of HEI, we observed a significant decreasing trend in the risk of PC (p for trend<0.001). Conclusion: Our findings suggested that a high quality diet, according to HEI, may decrease the risk of PC.


Introduction
Prostate cancer (PCa ) is the most diagnosed malignancy of men in Western countries (Grubb and Kibel, 2007). The incidence of PC in Asian countries, including Iran, is lower than that in Western populations . Various factors, including genetic factors, screening methods, lifestyle, and diet contribute to this difference (Baade et al., 2009;Zhang et al., 2011). Numerous studies focused on the effects of a nutrient on PCa. Some studies reported the effects of single food or nutrient on PCa (Giovannucci et al., 2002;Chavarro et al., 2007;Park et al., 2007). Dietary components are consumed in combination, affecting the absorption and bioavailability of each other (Michels and Schulze, 2005). Hence, identifying individual effect of dietary components is very difficult.
Dietary patterns allow to discover the link between the overall diet and the risk of disease (Moeller et al., 2007). Two of the most popular indices for measuring healthy eating patterns are The Healthy Eating Index (HEI) 2010 and The Mediterranean-Style Dietary Pattern Score (MSDPS). HEI 2010 examines the diet quality based on its conformance with Food Guide Pyramid and Dietary any changes in his diet since diagnosis; 4) be willing to participate in the study; 5) being a male. Based on our inclusion criteria, 100 men with PC were selected as case group (participation rate=71%). Based on inclusion criteria, 100 men with PC selected as case group. In addition, 205 patients who were admitted to the same hospitals (due to ear, nose, or eye disorder, appendicitis or general surgery) without any special diets were selected as the control group. Individuals in case and control groups were matched in terms of age (with a ten-year-interval). At the end, 3 patients in the case group whose their log scale of total energy intake were either >3SD or <3SD from the mean were excluded from the study. Therefore, 97 cases and 205 controls were underwent further analysis. All participants were interviewed to obtain needed information, including history of diabetes, hypertension, smoking status, alcohol use, and marital status. The weight of each participant with the least amount of clothing and a sensitivity of 100 gram was obtained by a digital scale, and the height of each of them was measured without shoes with a sensitivity of 0.1 cm. We measured waist circumference through this procedure: find the top of hip and the bottom of ribs, place the tap measure midway between these points, and wrap it around waist. The study was approved by Ethics Committee of Shahid Beheshti University of Medical Sciences.

Dietary assessment
Data on usual dietary intake of participants over the past year were collected through a valid and reliable semi-quantitative food frequency questionnaire (148-item FFQ) . All participants were asked to record food consumption frequency of every food item with standard serving size per day, per week, per month, and per year. The amount of each food item was convert to gram per day using the household scales (Ghaffarpour et al., 1999). Analyses of energy and nutrients were carried out using the USDA FCT (McGuire, 2011). However, for some dairy products such as Kashk, vetch, wild plum, mint, sweet canned cherry, and sour cherry that are not listed in the USDA FCT, Iranian FCT was used alternatively (Movahedi and Roosta, 2000). In order to calculate healthy eating components, we used national nutrient database to convert values obtained from FFQ to cup or ounce equivalents per 1000 kcal (Bowman et al., 2014).

Calculation of the Healthy Eating Index-2010
Healthy eating index-2010 (HEI-2010), as a measure of quality of life, was designed according to recommendations presented by the dietary guidelines for American 2010 (McGuire, 2011). The HEI is used to examine the relationships between diet and health-related outcomes. A summary of HEI-2010 components and their point values as well as its scoring standards are presented in Table 1. HEI-2010 includes 12 components divided into 9 adequacy components and 3 moderation components. Adequacy components consist of total fruit, whole fruit, total vegetables, greens and beans, whole grains, dairy, seafood and plant proteins, fatty acids, and total protein foods. Moderation components are refined grains, sodium, and empty calories (Guenther et al., 2013). All components scores were calculated based on their point values that is presented in Table 1. The participant with zero intake of adequacy components except fatty acids received a score of zero. For all adequacy components, the maximum number of points assigned to participants with intake at the level of the standard or higher. Other components scoring was carried out based on the standard for maximum score and minimum score of zero . Then, scores for the intake between the minimum and maximum standards were calculated proportionately.

Calculation of the Mediterranean-Style Dietary Pattern Score (MSDPS)
Mediterranean-Style Dietary Pattern Score (MSDPS) is consisted of 13 components based on Mediterranean pyramid. These 13 food groups are whole-grain cereals, fruits, vegetables, dairy products, wine, fish and other seafood, poultry, olives/legumes/nuts, potatoes and other starchy roots, eggs, sweets, meat, and olive oil (Rumawas et al., 2009a). In our study, only 12 components were available because participants did not respond to alcohol intake questions due to religious beliefs (Hosseini-Esfahani et al., 2010). With the exception of olive oil, each group was scored from 0 to 10 depending on the degree of correspondence with the recommendations. Olive oil was scored differently, score 10 was allocated to exclusive olive oil intake, score 5 to use of olive oil along with other vegetable oils, and score 0 to olive oil non-consumption . We also took into account the negative implications of overconsumption, defined as exceeding the recommended intake of foods in the Mediterranean diet pyramid. Overconsumption incurs a penalty by subtracting a point proportionally to the number of consumed servings that exceeded the recommended intake for that group. Finally for each participant, all 12 scores were summed up, and then were divided into proportion of total energy intake from Mediterranean diet in order to calculate the energy adjusted MSDPS that ranged from 0 (minimal adherence to Mediterranean pyramid recommendations) to 100 (maximal adherence) (Rumawas et al., 2009a).

Statistical analysis
Statistical analyses were conducted using SPSS version 19.0 (SPSS Inc., Chicago, IL, USA). Normal distribution of quantitative data was measured by the Kolmogorov-Smirnov test and normal distribution of qualitative data was measured by the Chi-square test. To compare differences between cases and controls, independent sample t-test and the Mann-whitney test were used for continuous variables and categorical variables, respectively.
Each of the density standards and the overall HEI score were categorized into tertiles based on the distribution among controls. We used binary logistic regression to estimate odds ratios and 95% confidence intervals for tertiles of each 12 component. Considered covariates included body mass index (continuous), waist circumstance, total energy intake (continuous), smoking status, and history of diabetes, and hypertension. Then, odds ratio was adjusted for this confounding factors.
(p<0.001), seafood, and plant protein(p=0.023). Sodium intake was significantly lower in the control group than that in the case group. Table 3 shows odds ratios (OR) and confidence intervals (95% CI) for PC by HEI-2010 components as well as total HEI-2010 scores. After adjusting the impact of confounders (energy intake, BMI, waist circumstance, family history of diabetes and blood pressure and smoking status). Individuals in the highest tertile of total fruit intake (OR=0.14; 95% CI=0.05-0.42), total vegetables (OR=0.23; 95% CI=0.09-0.59), whole fruit (OR=0.15; 95% CI=0.05-0.44), and fatty acids (OR=0.12; 95%CI=0.03-0.37) showed a significant decrease in the risk of PC compared to those in the lowest tertile of intake. There was no difference in terms of MSDPS between case and control groups (Table 5).

Results
Characteristics of 97 PC patients as cases and 205 controls are shown in Table 1. Cases and controls were not significantly different in terms of age, energy intake, alcohol use, and marital status, but they indicated significant differences in terms of BMI, waist circumference, history of diabetes, heart disease, hypertension, physical activity, and smoking status (P<0.05). Table 2 shows a comparison between the case (PC) and control groups based on the scores of HEI components and daily intake. Total HEI-2010 was significantly higher in the control group than the case group (p<0.001). The control group compared with the case group had higher intake and scores on components of whole fruit, total fruit, total vegetables, fatty acids Data are presented as frequency (percentage) or mean(SD); P-value was estimated using χ 2 or fisher test and t-test

Discussion
The associations between MSDPS and HEI 2010 and risk of PC were evaluated in this study. We found that participants with a higher adherence to HEI-2010 had a reduced risk of PC. In contrast, no significant association was found between MSDPS and risk of PC. Only few studies examined the index-based patterns and risk of PC. Bosire et al., (2013) found that high AHEI-2010 and HEI-2005 scores were associated with lower risk of total PC. The HEI-2010 seems to be more relevant than AHEI-2010 in decreasing cancer risk and mortality (Onvani et al., 2017). Differences between the two patterns may be due to this fact that HEI does not include alcoholic beverages and red and processed meat, whiles they are considered as AHEI components. According to previous studies, adherence to a Mediterranean dietary pattern was related to reduce risk of overall mortality and cancer incidence and mortality (Sofi et al., 2010;Verberne et al., 2010). However, several studies found no association between the Mediterranean diet score and PC (Tseng et al., 2004;Möller et al., 2013;Ax et al., 2014), our findings were consistent with those of other studies. The risk estimate is based on components of the Mediterranean diet pattern that is scored based on the intake of population distribution. The reported median intake in a Greek population is extremely higher than that in other populations listed here. The identified Mediterranean dietary pattern in these populations were not extensively Mediterranean, which may explain the different findings. Findings on dietary pattern and PC are inconsistent. A number of studies found no association between specific dietary patterns and PC (Tseng et al., 2004;Wu et al., 2006;Muller et al., 2009); however; the other studies reported a positive association between a western dietary pattern and PC . Individual component analyses were also performed, indicating significant inverse association between PC and seafood and plant protein as well as the ratio of total unsaturated/saturated fatty acids in HEI-2010 and the components of vegetables and fruits in both HEI-2010 and MED. Previous studies also found an inverse association between fish intake and PC risk (Augustsson et al., 2003;Chavarro et al., 2008;Pham et al., 2009). Augustsson et al., (Augustsson et al., 2003) showed that eating fish 3 or more times per week could reduce risk of metastatic PC 44% compared to its consumption less than 2 times per month. This association was also reported by Chavarro et al., (2008). A proposed mechanism for the protective effect of fish on PC is the long-chained ω-3 fatty acids present in fish. Both EPA and DHA have been found to inhibit the biological activity of eicosanoids and androgens (Faust et al., 1989;Zaccheo et al., 1998), which could stimulate the growth of PC cells (Ghosh and Myers, 1997;Rose, 1997 (Mills et al., 1989;Jain et al., 1999;Kolonel et al., 2000). The constituents of plant proteins that have anticarcinogenic properties and can be considered potentially protective against cancer could potentially account for a protective effect include fiber, protease inhibitors (Rao and Sung, 1995;Kennedy, 1998), saponins (Rao and Sung, 1995), inositol hexaphosphate ( et al., 1998), phytosterols (Steinmetz and Potter, 1996) and g-tocopherol (Steinmetz and Potter, 1996). We also found an association between other components and PC risk. The intake of fruits and vegetables, the components of both MED and HEI-2010, was inversely associated with PC risk. Previous reports also indicated that vegetables intake was associated with reduced PC risk (Kolonel et al., 2000;Takachi et al., 2009;Umesawa et al., 2014). There are several possible mechanisms for the inverse association between vegetable intake and risk of PC. For instance, vegetable components, including isothiocyanates and glucosinolates activate phase 2 enzyme, detoxificating carcinogen and stimulating cancer cell apoptosis (Hayes et al., 2008;Ho et al., 2009). The results on the association between fruit intake and PC risk are contrasting. In some studies, fruits intake was associated with a reduced risk of cancer (Mills et al., 1989). While other studies showed that fruit consumption did not have any effects on cancer risk (Severson et al., 1989;Schuurman et al., 1998). Evidence for the protective effect of fruits is limited and the reported findings are contradictory (Umesawa et al., 2014). Previous reports highlighted increased risk of PC following animal fat and saturated fat consumptions (Le Marchand et al., 1994;Pelser et al., 2013). In the current study, an inverse association between higher ratio of USFA/ SFA intake and PC risk was found. The prospective studies showed that SFA intake was related to the risk of advanced or fatal PC (Pelser et al., 2013). This increased effect can be due to the effect of SFA on the level of sex hormones. Hill et al., (1979) showed that reducing the consumption of animal fats and replacing them with vegetable oils led to lower levels of estrogens and androgens.
Case-control studies are subject to limitations that should be considered in interpreting their results. First, the possibility of selection bias cannot be avoided in retrospective case-control studies. In addition, our controls might also have nutritional problems, which could dilute the association of dietary intakes and risk of PC due to sharing of the exposure. However, we preferred hospital controls (opposed to community controls) due to their higher participation and cooperation rates and to avoid selection bias. The present study had several strengths. First, the participation rate in this study was high. Second, patients whose disease was diagnosed during previous 6 months were registered in order to reduce the possibility of recall bias. This is due to the fact that dietary data collected at the time of disease ascertainment might not truly reflect past intakes or intakes during the development of disease. Third, we used a valid FFQ to reduce measurement error.
In conclusion, the present study was designed to determine the effect of diet quality on PC risk. The result of this investigation showed that healthy diet based on dietary guidelines for American 2010 can reduce the risk of PC. These findings enhance our understanding of the effects of dietary quality on PC risk. Further investigation on dietary indices is strongly recommended. Considerably more work is needed to be done to determine the best dietary recommendation for cancer prevention.