Association between dietary phytoestrogens intakes and prostate cancer risk in Sicily

Abstract

Objective: In this study we aimed to investigate the association between dietary phytoestrogen consumption and prostate cancer in a sample of southern Italian individuals.

Methods: A population-based case-control study on the association between prostate cancer and dietary factors was conducted from January 2015 to December 2016 in a single institution of the municipality of Catania, southern Italy (Registration number: 41/2015). A total of 118 histopathological-verified prostate cancer (PCa) cases and a total of 222 controls were collected. Dietary data was collected by using two food frequency questionnaires.

Results: Patients with PCa consumed significantly higher levels of phytoestrogens. Multivariate logistic regression showed that lignans (Q[quartile]4 vs. Q1, OR [odds ratio] = 4.72; p < .05) and specifically, lariciresinol (Q4 vs. Q1, OR = 4.60; p < .05), pinoresinol (Q4 vs. Q1, OR = 5.62; p < .05), matairesinol (Q4 vs. Q1, OR = 3.63; p < .05), secoisolariciresinol (Q4 vs. Q1, OR = 4.10; p < .05) were associated with increased risk of PCa. Furthermore, we found that isoflavones (Q3 vs. Q1, OR = 0.28; p < .05) and specifically, genistein (Q4 vs. Q1, OR = 0.40; p < .05) were associated with reduced risk of PCa.

Conclusion: We found of an inverse association between dietary isoflavone intake and PCa, while a positive association was found with lignans intake.

Keywords:

• Phytoestrogens
• prostate cancer
• risk factors
• diet
• Mediterranean diet

Introduction

Prostate cancer (PCa) is the most common non-skin cancer and the second leading cause of cancer death among men in the United States [1–4]. Advanced age, race and a family history of PCa are established non-modifiable risk factors. In contrast, diet has been suggested to play a role in the pathogenesis of prostate cancer [5–10]. There is evidence that Western dietary patterns may increase the risk of PCa while adherence to plant-based dietary patterns may be associated with lower risk of prostate cancer [11]. However, the literature concerning a plant-based diet, related nutrients, phytochemicals and prostate cancer provided some contrasting results [12,13]. Overall, the evidence suggests that plant-based dietary patterns rich in fruits and vegetables may reduce risk of PCa [14]. Consuming a more plant based diet for cancer prevention is recommended by the World Cancer Research Fund/American Institute for Cancer Research and the American Cancer Society [15]. It has been hypothesized that most of the potential benefits related to plant-based dietary patterns may depend on specific nutrients, such as the carotenoids and vitamin C [16]. In addition, plant foods contain nutrient-like compounds (phytochemicals) that may have important anticarcinogenic activities that could potentially affect human health [17].

Furthermore, it has been previously reported that there is a coexistence of an altered hormone profile with increased sex hormones and leptin in PCa patients [3,18,19] and are markers of inflammation [20,21].

Among the most studied phytochemicals, phytoestrogens have been studied due to their structural and functional similarities to 17 b-oestradiol. Phytoestrogens have been shown to have estrogenic and antiestrogenic activity and the ability to modify other steroid hormones and hormone-binding proteins [22]. Dietary phytoestrogens are proposed to alter the risk of hormone-sensitive cancers primarily via competitive binding to estrogen receptors. However, the nature of the association between prostate cancer and estrogen exposure is yet to be determined and may depend on the timing of exposure [23]. The main studied classes of phytoestrogens in relation to cancer risk are isoflavones and lignans [24]. Thus our, research has mainly focused on the intake of isoflavones, a class of phytoestrogens contained at high concentrations in soy products, which are highly consumed in Asian countries. However, the phytoestrogen class lignans is widespread in a large range of foods, such as cereals, fruit, vegetables, grains, nut, and seeds [25]. Lower rates of hormone-sensitive cancers among Asian populations, characterized by regular consumption of phytoestrogen-rich foods, have contributed to speculation that phytoestrogen intake as part of a Western-style diet might similarly affect cancer risk [26]. However, population studies of the association between phytoestrogen intake and cancer risk among those who consume a Western-style diet have yielded inconclusive results [27]. Thus, the aim of this study was to investigate whether there was an association between dietary phytoestrogen consumption and prostate cancer in a sample of southern Italian individuals.

Material and methods

Study population

A population-based case-control study on the association between prostate cancer and dietary factors was conducted from January 2015 to December 2016 in a single institution of the municipality of Catania, southern Italy. Patients with elevated PSA and/or suspicious PCa underwent transperineal prostate biopsy (≥12 cores). A total of 118 histopathologically-verified PCa cases were collected.

Controls were selected from a sample of 2,044 individuals included in a cohort study by Grosso et al. [28]: individuals were randomly selected among the same reference population of the cases and were matched by age, BMI and smoking status with cases. A total of 222 controls were selected. All the study procedures were carried out in accordance with the Declaration of Helsinki (1989) of the World Medical Association and participants provided written informed consent after accepting to participate. The study protocol was approved by the ethic committee of the referent health authority.

Data collection

Demographics (including age and educational level) and lifestyle characteristics (including physical activity, smoking and drinking habits) were collected. Educational level was categorized as (i) low (primary/secondary), (ii) medium (high school) and (ii) high (university). Physical activity level was evaluated through the International Physical Activity Questionnaires (IPAQ) [29], which comprised a set of questionnaires (five domains) investigating the time spent being physically active in the last seven days: based on the IPAQ guidelines, final scores allowed to categorize physical activity level as (i) low, (ii) moderate and (iii) high. Smoking status was categorized as (i) non-smoker, (ii) ex-smoker and (iii) current smoker. Alcohol consumption was categorized as (i) none, (ii) moderate drinker (0.1–12 g/d) and (ii) regular drinker (>12 g/d).

Dietary assessment

Dietary data was collected by using two food frequency questionnaires (FFQs) specifically developed and validated for the Sicilian population [30,31]. The long-version FFQ consisted of 110 food and drink items that should refer to the participants’ diet during the last six months. Patients were specifically asked whether they changed their diet due to course of the disease and were asked to answer the questionnaire referring to their habitual diet before the disease. Participants were asked how often, on average, they had consumed foods and drinks included in the FFQ, with nine responses ranging from ‘never’ to ‘4–5 times per day’. Intake of food items characterized by seasonality referred to consumption during the period in which the food was available and then was adjusted by its proportional intake in one year.

Estimation of polyphenol intake

The methodology used to retrieve dietary polyphenols has been used in literature and largely described elsewhere [32]. Briefly, data on the polyphenol content in foods was obtained from the Phenol-Explorer database (www.phenol-explorer.eu). A new module of the Phenol-Explorer database containing information on the effects of cooking and food processing on polyphenol contents was used whenever possible in order to apply polyphenol-specific retention factor. A total of 75 items were searched in the database after exclusion of foods that contained no polyphenols. Following the standard portion sizes used in the study, food items were converted in g or ml and then proportioned to 24 h intake. Then, a search was carried out in the Phenol-Explorer database to retrieve mean content values for phytoestrogens (major subclasses and selected compounds) contained in the foods obtained and their intake was then calculated by multiplying the phytoestrogen content by the daily consumption of each food. Finally, intake of phytoestrogens was adjusted for total energy intake (kcal/d) using the residual method.

Statistical analysis

Categorical variables are presented as frequencies and percentages, continuous variables are presented as means and standard deviations. Differences of frequency between groups were calculated by Chi-square test. Phytoestrogen intake distribution was tested for normality distribution with the Kolmogorov-Smirnov test and it followed a slightly asymmetric normal distribution due to extreme values of the upper side. Mann-Whitney U-test and Kruskall-Wallis test were used to compare differences in intakes between groups, as appropriate. Association between dietary intake of phytoestrogens and PCa was calculated through logistic regression analysis adjusted for age (years, continuous), energy intake (kcal/d, continuous), weight status (normal, overweight, obese), smoking status (smokers, non-smokers), alcohol consumption (<12 g/d, ≥12 g/d), physical activity level (low, medium, high), family history of PCa. All reported p values were based on two-sided tests and compared to a significance level of 5%. SPSS 17 (SPSS Inc., Chicago, IL) software was used for all the statistical calculations.

Results

Table 1 lists the baseline characteristics of the patients. No significant differences between cases and controls were retrieved, with the exception for family history of cancer, which was more common among cases and education, which was higher among controls.

Table 1. Baseline characteristics of cases and controls.

	Cases (n = 118)	Controls (n = 222)	p value
Age (y), mean (SD)	69.13 (6.60)	68.09 (8.18)	.19
BMI, mean (SD)	26.49 (3.34)	27.49 (3.28)	.30
Weight status, n (%)
Normal	42 (35.6)	59 (26.6)
Overweight	60 (50.8)	127 (57.2)
Obese	16 (13.6)	36 (16.2)
Smoking status, n (%)			.15
Non-smoker	68 (57.6)	143 (64.4)
Current smoker	50 (42.4)	79 (35.6)
Alcohol intake, n (%)			.16
<12 g/d	55 (46.6)	153 (68.9)
≥12 g/d	63 (53.4)	69 (31.1)
Education, n (%)			.11
Primary/secondary	96 (81.4)	49 (22.1)
High school/university	22 (18.6)	173 (77.9)
Physical activity level, n (%)			.21
Low	38 (32.2)	49 (26.2)
Medium	64 (54.2)	67 (35.8)
High	16 (13.6)	71 (38.0)
Family history of prostatic cancer, n (%)	43 (36.44)	9 (4.05)	<0.01

Mean consumption of dietary phytoestrogens is presented in Table 2. Patients with PCa consumed significantly higher levels of lignans (7.00 mg/d vs. 3.11 mg/d; p < .01), lariciresinol (4.04 mg/d vs. 1.68 mg/d; p < .01), pinoresinol (2.28 mg/d vs. 1.06 mg/d; p < .01) and secoisolariciresinol (0.26 mg/d vs. 0.12 mg/d; p < .01) than controls.

Table 2. Differences in mean consumption of phytoestrogens between cases and controls.

	Cases (n = 118)	Controls (n = 222)	p value
Lignans, mean (SD)	7.00 (5.43)	3.11 (2.70)	<.01
Isoflavones, mean (SD)	2.51 (10.51)	1.93 (8.75)	.59
Lariciresinol, mean (SD)	4.04 (3.29)	1.68 (1.64)	<.01
Matairesinol, mean (SD)	0.08 (0.06)	0.03 (0.03)	<.01
Pinoresinol, mean (SD)	2.28 (1.69)	1.06 (0.84)	<.01
Secoisolariciresinol, mean (SD)	0.26 (1.19)	0.12 (1.10)	<.01
Daidzein, mean (SD)	0.08 (0.24)	0.09 (0.33)	.88
Genistein, mean (SD)	0.08 (0.28)	0.08 (0.32)	.93
Glycitein, mean (SD)	0.02 (0.08)	0.01 (0.07)	.62
Biochanin A, mean (SD)	0.01 (0.002)	0.01 (0.003)	.32

Table 3. Association between quartiles of phytoestrogen intake and prostate cancer.

	Phytoestrogens quartiles, OR (95% CI)
	Q1	Q2	Q3	Q4
Lignans
No. of cases	14	8	10	86
OR (95% CI)^a	Ref.	0.51 (0.19–1.31)	0.57 (0.23–1.45)	4.72 (2.34–9.52)
OR (95% CI)^b	Ref.	0.37 (0.10–1.31)	0.70 (0.17–2.81)	9.31 (3.16–27.38)
Isoflavones
No. of cases	39	38	13	28
OR (95% CI)^a	Ref.	0.80 (0.44–1.45)	0.31 (0.15–0.65)	0.51 (0.27–0.96)
OR (95% CI)^b	Ref.	1.40 (0.58–3.40)	0.28 (0.10–0.77)	0.95 (0.39–2.33)
Lariciresinol
No. of cases	15	9	6	88
OR (95% CI)^a	Ref.	0.56 (0.22–1.41)	0.32 (0.11–0.93)	4.60 (2.32–9.11)
OR (95% CI)^b	Ref.	0.27 (0.08–0.98)	0.39 (0.08–2.00)	8.42 (2.94–24.13)
Matairesinol
No. of cases	17	7	16	78
OR (95% CI)^a	Ref.	0.38 (0.15–1.01)	0.80 (0.35–1.77)	3.63 (1.86–7.10)
OR (95% CI)^b	Ref.	0.31 (0.09–1.11)	1.02 (0.30–3.40)	6.49 (2.35–17.89)
Pinoresinol
No. of cases	12	8	13	85
OR (95% CI)^a	Ref.	0.60 (0.23–1.60)	0.98 (0.40–2.40)	5.62 (2.70–11.70)
OR (95% CI)^b	Ref.	0.54 (0.14–2.08)	1.45 (0.39–5.46)	12.77 (4.14–39.36)
Secoisolariciresinol
No. of cases	16	8	10	84
OR (95% CI)^a	Ref.	0.44 (0.17–1.12)	0.50 (0.20–1.23)	4.10 (2.10–8.10)
OR (95% CI)^b	Ref.	0.23 (0.06–0.81)	0.50 (0.13–1.96)	6.30 (2.25–17.60)
Daidzein
No. of cases	38	39	11	30
OR (95% CI)^a	Ref.	0.81 (0.45–1.46)	0.39 (0.18–0.87)	0.45 (0.24–0.83)
OR (95% CI)^b	Ref.	1.22 (0.51–2.92)	0.32 (0.11–0.94)	0.71 (0.30–1.67)
Genistein
No. of cases	35	46	12	25
OR (95% CI)^a	Ref.	0.97 (0.54–1.75)	0.42 (0.19–0.93)	0.40 (0.21–0.77)
OR (95% CI)^b	Ref.	1.43 (0.60–3.40)	0.33 (0.11–0.97)	0.60 (0.25–1.48)
Glycitein
No. of cases	110	8
OR (95% CI)^a^,c	Ref.	0.52 (0.21–1.31)	–	–
OR (95% CI)^b^,c	Ref.	0.46 (0.15–1.48)	–	–
Biochanin A
No. of cases	42	20	56
OR (95% CI)^a^,d	Ref.	0.59 (0.32–1.10)	1.66 (0.97–2.83)	–
OR (95% CI)^b^,d	Ref.	0.40 (0.16–0.99)	1.39 (0.63–3.08)	–

Q: quartile; OR: odds ratio.

^a ORs were adjusted for energy intake (kcal/d, continuous).

^b ORs were adjusted for age (years, continuous), energy intake (kcal/d, continuous), weight status (normal, overweight, obese), smoking status (smokers, non-smokers), alcohol consumption (<12 g/d, ≥12 g/d), physical activity level (low, medium, high) and family history of prostatic cancer;.

^c Due to the low number of cases and low amount of compound consumed, individuals were divided into two quantiles (less and more than median intake).

^d Due to the low number of cases and low amount of compound consumed, individuals were divided into tertiles.

Table 4. Association between quartiles of phytoestrogen intake and advanced prostate cancer.

	Phytoestrogens quartiles, OR (95% CI)^a
	Q1	Q2	Q3	Q4
Isoflavones	Ref.	0.47 (0.13–1.73)	0.13 (0.15–1.09)	0.01 (0.00–1.97)
Lignans^a	Ref.	0.97 (0.15–6.26)	–	–
Lariciresinol^b	Ref.	1.08 (0.17–6.99)	–	–
Matairesinol^b	Ref.	0.90 (0.13–6.14)	–	–
Pinoresinol^c	Ref.	0.48 (0.03–8.47)	2.47 (0.25–23.90)	–
Secoisolariciresinol^b	Ref.	0.72 (0.09–5.29)	–	–
Daidzein	Ref.	0.43 (0.11–1.58)	0.17 (0.02–1.49)	0.01 (0.00–1.81)
Genistein	Ref.	0.66 (0.18–2.32)	0.20 (0.02–1.74)	0.01 (0.00–1.75)
Biochanin A	Ref.	2.50 (0.22–28.22)	12.46 (1.52–101.87)	1.00 (1.00–1.00)

Q: quartile; OR: odds ratio.

^a ORs were adjusted for energy intake (kcal/d, continuous).

^b Due to the low number of cases and low amount of compound consumed, individuals were divided into two quantiles (less and more than median intake).

^c Due to the low number of cases and low amount of compound consumed, individuals were divided into tertiles.

Multivariate logistic regression adjusted for age, energy intake, weight status, smoking status, alcohol consumption, physical activity level and family history of prostatic cancer showed that lignans (Q[quartile]4 vs. Q1, OR[odds ratio] = 4.72 [95% CI: 2.34–9.52]) and specifically, lariciresinol (Q4 vs. Q1, OR = 4.60 [95% CI: 2.32–9.11]), pinoresinol (Q4 vs. Q1, OR = 5.62 [95% CI: 2.70–11.70]), matairesinol (Q4 vs. Q1, OR = 3.63 [95% CI: 1.86–7.10]) and secoisolariciresinol (Q4 vs. Q1, OR = 4.10 [95% CI: 2.10–8.10]) were associated with increased risk of PCa (Table 3). Furthermore, we found that isoflavones (Q3 vs. Q1, OR = 0.28 [95% CI: 0.10–0.77]) and specifically, genistein (Q4 vs. Q1, OR = 0.40 [95% CI: 0.21–0.77]) were associated with reduced risk of PCa. We found no other significant association between different phytoestrogen and advanced PCa risk (Table 4).

Discussion

In this study we investigated the association between phytoestrogen intake and PCa incidence. We found that the association between phytoestrogens and PCa was different according to the subclass of compound analyzed. Specifically, we found that high intake of lignans, lariciresinol, pinoresinol, matairesinol and secoisolariciresinol were positively associated with PCa. On the contrary, isoflavones and genistein were significantly inversely associated with PCa.

The relation between dietary factors and PCa has been investigated and one explanation for the low incidence of the cancer in Asia has been hypothesized to be related to the high consumption of soybeans and their derived products [33–35], which are rich in isoflavones. In Western populations, characterized by a low intake of isoflavones, phytoestrogen intake is predominantly derived from intake of plant lignans. Based on previous publication, the sample selected as control in this study had a generally high adherence to the Mediterranean diet [32]; major sources of isoflavones have been reported to be soy foods (mainly milk) and major sources of lignans were citrus fruit. Consequently, it is difficult to say that consumption of such compounds was associated with foods potentially increasing the risk of PCa.

Previous studies investigating phytoestrogen intake and prostate cancer have been limited to soy foods as a measure of phytoestrogen intake [36]. A recent meta-analysis to address this gap updated and assessed quantitatively the association between intake of isoflavones and lignans and risk of prostate cancer from the cohort and case-control studies and investigated the association between serum concentration of phytoestrogens and their metabolites and the risk of PCa [37]. Their findings support the hypotheses that consumption of genistein and daidzein may protect against PCa risk. However, the complexity of phytoestrogen composition and its metabolism make the evaluation of the effect of phytoestrogen on PCa very difficult. In vivo and in vitro experiments have also shown a protective effect of isoflavones against prostate cancer development [38]. Isoflavones possess weak estrogen activity, inhibiting tyrosine protein kinases and angiogenesis and reduce serum testosterone levels; isoflavones also inhibit 5 alpha-reductase, an enzyme that metabolizes testosterone to dihydrotestosterone [39]. One mechanism by which isoflavones reduce the risk of prostate cancer seems to involve estrogen receptor H in prostate tissue [40], but cancer with higher metastatic potential is associated with the complete or partial loss of estrogen receptor H expression [41]. On this basis, isoflavones may prevent the early stages of prostate cancer development but fail in preventing at the advanced stages.

Studies on dietary lignans and PCa risk are scarce. Previous investigations showed overall contrasting results on the association between lignans, its metabolites and PCa [24]. In our study we found that lignan intake was positively associated with PCa. A previous study showed an increased risk of PCa associated with total and individual lignan intake, despite the association was no longer significant after adjusting for other dietary factors (dairy foods) [42]. We are unaware whether in our case-control study such confounding effect of other dietary factors may exist. However, consumption of lignans in our sample was generally related to plant foods, which should protect, or at least have null effect towards PCa risk. Besides the estrogenic activity, lignans have demonstrated anticancer effects through inhibition of proliferation and apoptosis promotion. However, their hormonal-related activity may somehow alter the physiological hormonal homeostasis leading to increased exposure to pro-estrogenic action and current research is ongoing to better understand the dualistic mode of action in relation to cell proliferation and cancer risk.

The results of the present study should be considered in light of some limitations. First, the FFQs used to assess dietary habits may lead to measurement errors not just due to recall bias but also due to the estimation by using different food composition databases, which may not be complete for the whole range of foods consumed. Phytoestrogen spectrum and content varies between the plant species, sort and origin; even the same molecule arising from the different sources can exert various effects. Moreover, equol is a metabolite of daidzein produced by the intestinal microflora that has higher estrogenic activity than its parent isoflavone. The most abundant lignan in human subjects is enterolactone, which is produced by certain types of intestinal microflora from plant lignan glycosides. Variation in individual metabolism of phytoestrogens due to differences in gut microflora may influence the serum concentration of phytoestrogens and their biological effects [43]. It has been reported that the capacity to produce equol was lower among American than Japanese and Korean men [44]. Thus, dietary intake may not correspond to clinically relevant absorbed compound and metabolites concentrations. Second, a common limitation of population-based case-control studies including non-screened population as controls is that we are unaware whether some undiagnosed PCa occurred in the control group. However, the rate of potentially undiagnosed cancer would be low and is not likely to affect the results of the study.

Conclusion

In conclusion, we provided further evidence of an inverse association between dietary isoflavone intake and PCa, while a positive association was found with lignans intake. In the light of these findings, further prospective epidemiological studies using improved food databases and markers of consumption are needed to identify the specific compounds that provide protection, in order to determine precisely how the complex metabolism of phytoestrogens may interact with other mechanisms to prevent cancer.

Disclosure statement

No potential conflict of interest was reported by the authors.