Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2018 Sep 29;143(11):2677–2686. doi: 10.1002/ijc.31640

Circulating isoflavone and lignan concentrations and prostate cancer risk: a meta‐analysis of individual participant data from seven prospective studies including 2,828 cases and 5,593 controls

Aurora Perez‐Cornago 1,✉, Paul N Appleby 1, Heiner Boeing 2, Leire Gil 3,4, Cecilie Kyrø 5, Fulvio Ricceri 6,7, Neil Murphy 8, Antonia Trichopoulou 9, Konstantinos K Tsilidis 10,11, Kay‐Tee Khaw 12, Robert N Luben 12, Randi E Gislefoss 13, Hilde Langseth 13, Isabel Drake 14, Emily Sonestedt 14, Peter Wallström 14,15, Pär Stattin 16, Anders Johansson 17, Rikard Landberg 18,19, Lena Maria Nilsson 19,20, Kotaro Ozasa 21, Akiko Tamakoshi 22, Kazuya Mikami 23, Tatsuhiko Kubo 24, Norie Sawada 25, Shoichiro Tsugane 25, Timothy J Key 1, Naomi E Allen 26,27, Ruth C Travis 1
PMCID: PMC6283047  PMID: 29971774

Abstract

Phytoestrogens may influence prostate cancer development. This study aimed to examine the association between prediagnostic circulating concentrations of isoflavones (genistein, daidzein, equol) and lignans (enterolactone and enterodiol) and the risk of prostate cancer. Individual participant data were available from seven prospective studies (two studies from Japan with 241 cases and 503 controls and five studies from Europe with 2,828 cases and 5,593 controls). Because of the large difference in circulating isoflavone concentrations between Japan and Europe, analyses of the associations of isoflavone concentrations and prostate cancer risk were evaluated separately. Prostate cancer risk by study‐specific fourths of circulating concentrations of each phytoestrogen was estimated using multivariable‐adjusted conditional logistic regression. In men from Japan, those with high compared to low circulating equol concentrations had a lower risk of prostate cancer (multivariable‐adjusted OR for upper quartile [Q4] vs. Q1 = 0.61, 95% confidence interval [CI] = 0.39–0.97), although there was no significant trend (OR per 75 percentile increase = 0.69, 95 CI = 0.46–1.05, p trend = 0.085); Genistein and daidzein concentrations were not significantly associated with risk (ORs for Q4 vs. Q1 = 0.70, 0.45–1.10 and 0.71, 0.45–1.12, respectively). In men from Europe, circulating concentrations of genistein, daidzein and equol were not associated with risk. Circulating lignan concentrations were not associated with the risk of prostate cancer, overall or by disease aggressiveness or time to diagnosis. There was no strong evidence that prediagnostic circulating concentrations of isoflavones or lignans are associated with prostate cancer risk, although further research is warranted in populations where isoflavone intakes are high.

Keywords: prostate cancer risk, phytoestrogens, isoflavones, lignans, pooled analysis

Short abstract

What's new?

The role of phytoestrogens in prostate cancer development is uncertain. Here, the authors analysed participant data from seven prospective studies on the association between pre‐diagnostic circulating concentrations of isoflavones (mainly found in soybeans) and lignans (mainly found in cereal, nuts, and vegetables) and prostate cancer risk. They found no strong associations but point to the fact that further data are needed to examine associations based on disease aggressiveness, especially in populations with high isoflavone intakes.

Abbreviations

BMI

body mass index

CI

confidence interval

OR

odds ratio

1. Introduction

It has been hypothesised that a high dietary phytoestrogen intake may inhibit prostate cancer development.1, 2 Evidence from in vitro studies, animal models and observational studies has suggested that this may be due to their effects on hormone metabolism and receptor signalling, or to other mechanisms such as effects on DNA repair, angiogenesis and cell proliferation.3. The main types of dietary phytoestrogens are isoflavones and lignans. Soybeans and soya products are the predominant dietary source of isoflavones,4 while precursors of mammalian lignans are mainly present in cereals, nuts, seeds and fruits and vegetables.5, 6 The absorption of phytoestrogens is influenced by the gut microbiota,7, 8 therefore it is important to measure these compounds in biological samples such as blood (rather than relying solely on an estimation of dietary intake) in order to assess the possible protective effect of phytoestrogens on prostate cancer development.

The Endogenous Hormones, Nutritional Biomarkers and Prostate Cancer Collaborative Group (EHNBPCCG) was established to conduct collaborative analyses of individual participant data from prospective studies on the associations of circulating concentrations of hormones and nutritional biomarkers with risk of prostate cancer.9, 10, 11, 12, 13, 14 The objective of the present study was to examine the associations between concentrations of circulating isoflavones (genistein, daidzein and equol) and lignans (enterolactone and enterodiol), and risk of prostate cancer in a pooled analysis of the worldwide prospective studies.

2. Material and Methods

2.1. Identification of studies

Studies were eligible to join this collaborative analysis if they had data on circulating phytoestrogens measured in blood samples collected prior to a prostate cancer diagnosis. Studies were identified by using the search terms phyto‐oestrogens, phytoestrogens, isoflavones, genistein, daidzein, equol, enterolactone and enterodiol together with the MeSH term prostatic neoplasms and text term prostate cancer. The literature searches were conducted using articles with publication dates through to November 2017 from PubMed, Web of Science, Cochrane Library, CancerLit (up to January 2013) and Google Scholar and from discussions with colleagues. This collaborative group also includes unpublished data obtained from discussion with colleagues, which reduces the risk of publication bias.

2.2. Collection and synthesis of data

Individual participant data were available from seven nested case–control studies by the date of dataset closure (November 2017): the European Prospective Investigation into Cancer and Nutrition (EPIC) (cases diagnosed from June 1999 to January 2003 (designated Phase 1);15 cases diagnosed from February 2003 to December 2006 (designated Phase 2),16 EPIC‐Norfolk17 (for participants not included in either the EPIC Phase 1 or Phase 2), Janus Nordic Biological Specimen Biobank Working Group (NBSBWG),18, 19, 20 Japan Collaborative Cohort Study (JACC),21 Japan Public Health Center‐based prospective Study (JPHC),22 the Malmö Diet and Cancer Study (MDCS)23 and the Northern Sweden Health and Disease Cohort (NSHDC).24 Three of these studies contributed data on genistein (1,846 cases, 2,200 controls), four had data on daidzein (1,239 cases, 1,675 controls) and equol (1,209 cases, 1,571 controls), five had data on enterolactone (2,828 cases, 5,593 controls) and two studies had data on enterodiol (1,002 cases, 1,197 controls). Data on enterolactone from two studies (the α‐Tocopherol, β‐Carotene Cancer Prevention Study and the Helsinki Heart Study), with a combined total of 350 cases, were not available for pooling.18, 25

This large collaborative analysis has brought together and reanalysed almost all of the available prospective data on the associations of circulating isoflavone and lignan concentrations with prostate cancer incidence, representing, as far as we are aware, all of the worldwide prospective data on circulating genistein, daidzein and equol, and over 90% of the worldwide data on circulating lignans. While data on enterolactone were not available for this analysis, their results do not differ materially from those reported here, and it is unlikely that these data would have changed our summary risk estimates.

The characteristics of the studies and the assay methods are shown in Tables S1 and S2; detailed information on recruitment, ethics approval and inclusion criteria is available in the original publications.15, 16, 17, 18, 21, 22, 24

Where available, collaborators provided data for prostate cancer cases and controls on serum or plasma phytoestrogen concentrations, dates of birth and blood collection, marital status, ethnicity, educational attainment, family history of prostate cancer, height, weight and smoking status. Most studies also provided data on cancer stage and grade and date at diagnosis. Tumour stage was categorised as localised if it was tumour‐node‐metastasis (TNM) stage ≤T2 with no reported lymph node involvement or metastases, stage ≤II, or equivalent; advanced stage if it was T3 or T4 and/or N1+ and/or M1, stage III‐IV, or equivalent; or unknown. Aggressive disease was categorised as “no” for TNM stage ≤T3 with no reported lymph node involvement or metastases or the equivalent, “yes” for TNM stage T4 and/or N1+ and/or M1 and/or stage IV disease or death from prostate cancer, or unknown. Histological grade was categorised as low‐intermediate grade if the Gleason sum was <8, or coded as well, moderately or poorly differentiated; high grade if the Gleason sum was ≥8, or coded as undifferentiated; or grade unknown.

2.3. Statistical analyses

The methods of analysis were similar to those described previously by this collaborative group.10, 11 Because of the large difference in circulating isoflavone concentrations between Asian and Western populations, the analyses for isoflavones were performed separately for Japanese and European studies.

Men were categorised into fourths of concentrations for each phytoestrogen, with cut points defined by the study‐specific quartiles of the distribution in control participants. This approach allows for any systematic differences between the studies, i.e., assay methods, blood sample types or storage conditions.26 We used logistic regression conditioned on the matching variables within each study to determine odds ratios (ORs) and their corresponding 95% confidence intervals (CIs). To provide a summary measure of the OR (for subgroup analyses) and to calculate a p‐value for trend, the categorical variable representing the fourths of phytoestrogen concentration was replaced with a continuous variable that was scored 0, 0.33, 0.67 and 1, respectively. Because the mid‐points of the lowest and highest fourths are the 12.5 and 87.5 percentage points of the study‐specific phytoestrogen concentration, a unit increase in this variable can be taken to represent a 75 percentile increase in phytoestrogen concentration. The effects of additional potential confounders (other than the matching criteria, which were controlled by study design) were examined by additionally adjusting for variables that were found to be associated with prostate cancer risk in this collaborative dataset,10, including age at blood collection (exact), body mass index (BMI = <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown) and cigarette smoking (never, past, current, unknown).

Heterogeneity in linear trends between studies was tested by comparing the chi‐squared values for models with and without a (study) × (linear trend) interaction term. To identify evidence of heterogeneity in the associations of phytoestrogens with prostate cancer risk according to case‐defined characteristics (i.e., age at diagnosis [<65 vs. ≥ 65 years], years between blood collection and diagnosis [<5 vs. ≥ 5 years], year of diagnosis [pre‐2000 vs. 2000 onwards], stage [localised vs. advanced], aggressiveness [no vs. yes] and grade [low‐intermediate vs. high]), we fitted separate models for each subgroup and assumed independence of the ORs using a method analogous to a meta‐analysis. Data on tumour stage and/or grade were only available for one of the two Japanese studies,22 and hence subgroup analyses of the association of isoflavone concentrations by tumour characteristics in these studies were not performed.

Tests for heterogeneity for non‐case‐defined factors (i.e., age at blood collection [<65 vs. ≥ 65 years], BMI [<25 vs. ≥ 25 kg/m2], cigarette smoking [never or past smoker vs. current smoker] and alcohol consumption [<10 vs. ≥ 10 g/day]) were assessed with chi‐squared tests of interaction between the subgroup and the continuous trend test variable.

All tests of statistical significance were two sided, and statistical significance was set at the 5% level. All statistical analyses were carried out using Stata version 14.1 (Stata Corp., College Station, TX).

3. Results

A total of 241 cases and 503 controls from two Japanese prospective studies, and 2,828 cases and 5,593 controls from five European prospective studies were included in this analysis. The mean age at blood collection in cases or controls across the studies ranged from 46.5 to 68.7 years (Table 1).

Table 1.

Participant characteristics by study and case–control status

Studies Case–control status Participants, n Age at recruitment, years Height, 1 cm BMI, 1 kg/m2 Married or Cohabiting, 1 % Higher Education, 1 % Current Smoker, 1 %
Japanese
JACC (22) Case 40 68.7 (6.3) 160 (6) 22.4 (2.6) 100 9.7 51.4
Control 101 67.9 (5.7) 159 (7) 22.4 (2.7) 92.3 26.4 37.8
JPHC (23) Case 201 59.5 (6.4) 162 (6) 23.4 (2.4) 93.5 – 34.3
Control 402 59.2 (6.6) 162 (6) 23.3 (2.6) 90.5 – 40.8
European
EPIC, Phase 1 (16) Case 950 59.9 (5.8) 174 (7) 26.6 (3.4) 89.2 26.9 24.3
Control 1,042 59.6 (5.8) 174 (7) 26.9 (3.6) 88.9 26.4 29.8
EPIC, Phase 2 (17) Case 655 59.3 (6.9) 173 (7) 26.9 (3.4) 89.5 25.7 21.5
Control 655 59.3 (6.9) 173 (7) 26.9 (3.6) 90.2 26.4 23.1
EPIC‐Norfolk (18) Case 48 65.3 (6.7) 172 (8) 26.3 (2.7) 87.0 12.5 10.4
Control 130 64.7 (6.4) 173 (6) 26.5 (2.9) 90.8 11.5 10.2
Janus NBSBWG (19) Case 573 46.5 (4.3) 177 (7) 25.4 (3.1) – – 60.9
Control 2,209 46.5 (4.2) 176 (7) 25.1 (3.2) – – 62.4
MDCS (24) Case 990 60.2 (6.6) 177 (7) 26.2 (3.4) 78.5 15.1 22.3
Control 1,664 60.0 (6.6) 176 (6) 26.3 (3.4) 74.0 13.5 27.1
NSHDC (25) Case 261 58.0 (4.4) 176 (6) 26.1 (2.8) 87.1 13.2 18.2
Control 514 58.0 (4.4) 175 (6) 26.6 (3.5) 79.8 12.2 22.1
Open in a new tab

The cases and controls are nested in prospective studies, and the numbers of cases and controls are based on the number in complete matched sets for genistein (EPIC, JACC and JPHC), daidzein (EPIC‐Norfolk) and enterolactone (Janus NBSBWG, MDCS and NSHDC).

1

Unknown for some participants.

Abbreviations: BMI, body mass index; EPIC, European Prospective Investigation into Cancer and Nutrition; NBSBWG, Janus Nordic Biological Specimen Biobank Working Group; JACC, Japan Collaborative Cohort Study; JPHC, Japan Public Health Center‐based prospective Study; MDCS, Malmö Diet and Cancer Study; NSHDC, Northern Sweden Health and Disease Cohort.

Most of the cases were between 60 and 69 years at diagnosis, except for those from Japanese studies and EPIC‐Norfolk where most cases were diagnosed over 70 years (Table 2). The majority of cases with information on stage and grade had localised disease (ranging from 70 to 87% across studies) and low‐intermediate grade tumours (75 to 98% of cases).

Table 2.

Characteristics of men who developed prostate cancer and tumour characteristics by study

Age at diagnosis, % Date of diagnosis, % Years from blood collection to diagnosis, % Disease stage, aggressiveness and grade, %
Studies <60 60–69 ≥70 Pre‐2000 2000‐ <3 3–6 ≥7 Advanced stage 1 Unknown stage Aggressive disease 2 Unknown aggressiveness High grade 3 Unknown grade
Japanese
JACC (22) 0 35.0 65.0 100 0 17.5 52.5 30.0 – 100 – 100 – 100
JPHC (23) 7.0 39.8 53.2 25.4 74.6 9.0 17.4 73.6 25.6 24.9 29.9 23.4 24.2 69.1
European
EPIC, Phase 1 (16) 19.4 65.0 15.6 46.7 53.3 31.0 55.7 13.3 28.9 27.2 31.5 21.6 13.6 22.6
EPIC, Phase 2 (17) 14.2 55.0 30.8 3.8 98.2 3.0 42.9 54.0 24.8 28.5 19.5 26.4 11.6 25.0
EPIC‐Norfolk (18) 4.2 33.3 62.5 2.1 97.9 2.1 62.5 35.4 17.8 6.2 31.1 6.2 25.0 91.7
Janus NBSBWG (19) 20.9 69.1 9.9 100 0 1.2 5.1 93.7 – 100 – 100 – 100
MDCS (24) 5.3 48.7 46.1 25.1 74.9 9.5 25.3 65.3 – 100 1004 89.7 – 100
NSHDC (25) 16.9 78.5 4.6 72.0 28.0 27.6 49.8 22.6 18.5 0.8 12.7 0.8 1.5 74.3
Open in a new tab

The percentages are based on the number in complete matched sets for genistein (EPIC, JACC and JPHC), daidzein (EPIC‐Norfolk), and enterolactone (Janus NBSBWG, MDCS and NSHDC).

1

As a percentage of those with known disease stage. Stage of disease was defined as being advanced if it was tumour‐node‐metastasis (TNM) stage T3 or T4 and/or N1+ and/or M1, stages III–IV, or approximate equivalent (that is, a tumour extending beyond the prostate capsule and/or lymph node involvement and/or distant metastases), localised if it was TNM stage T0 or T1 or T2 with no reported lymph node involvement or metastases, stages 0–II, or approximate equivalent (that is, a tumour that does not extend beyond the prostate capsule) or stage unknown.

2

As a percentage of those with known aggressive disease. Aggressive disease was categorised as “yes” for TNM stage T4 and/or N1+ and/or M1 and/or stage IV disease or death from prostate cancer, “no” for TNM stage T0, T1, T2 or T3 with no reported lymph node involvement or metastases or the equivalent, or unknown.

3

As a percentage of those with known disease grade. Grade of disease defined as high grade if the Gleason sum was at least 8 or approximate equivalent (that is, extent of differentiation of “none”), low grade if the Gleason sum was less than 8 or approximate equivalent (that is, extent of differentiation of “poor,” “moderate” or “good”) or grade unknown.

4

Stage and grade of disease are unknown in MDCS, so aggressive disease is defined as cases who died of prostate cancer, but we cannot determine whether cases who did not die of prostate cancer had aggressive disease. Therefore, where aggressive disease status is known for this study, aggressive disease status is ‘yes’ and cannot be ‘no.’

Abbreviations: EPIC, European Prospective Investigation into Cancer and Nutrition; NBSBWG, Janus Nordic Biological Specimen Biobank Working Group; JACC, Japan Collaborative Cohort Study; JPHC, Japan Public Health Center‐based prospective Study; MDCS, Malmö Diet and Cancer Study; NSHDC, Northern Sweden Health and Disease Cohort.

The geometric means and interquartile ranges of the five phytoestrogens by case–control status are shown in Table 3. There was a large difference in circulating isoflavone concentrations between Japanese and European populations; for example, the mean genistein concentration in controls was between 294.0 and 454.4 nmol/L in Japanese studies and between 5.19 and 5.61 nmol/L in European studies. There was no large variation in circulating lignan concentrations between studies.

Table 3.

Prediagnostic geometric mean (95% CI) phytoestrogen concentrations (nmol/L) by study in cases and controls

Lignans
Studies Isoflavones Genistein Daidzein Equol Enterolactone Enterodiol
Japanese
JACC (22) Case 331.8 (250.6–439.2) 127.6 (92.8–175.3) 10.3 (5.6–19.1) – –
Control 454.4 (361.3–571.6) 166.5 (130.5–212.3) 24.0 (16.2–35.7) – –
JPHC (23) Case 277.2 (235.5–326.2) 115.9 (96.6–139.0) 13.6 (10.6–17.4) – –
Control 294.0 (261.7–330.2) 122.8 (107.7–140.0) 17.4 (14.5–20.9) – –
European
EPIC, Phase 1 (16) Case 4.84 (4.29–5.48) 3.80 (3.42–4.22) 0.65 (0.61–0.70) 11.0 (10.1–11.9) 0.99 (0.91–1.07)
Control 5.61 (5.00–6.29) 3.96 (3.58–4.38) 0.65 (0.60–0.69) 11.5 (10.6–12.4) 0.99 (0.92–1.08)
EPIC, Phase 2 (17) Case 5.97 (5.26–6.79) – – – –
Control 5.19 (4.59–5.87) – – – –
EPIC‐Norfolk (18) Case – 3.31 (2.08–5.27) 0.59 (0.28–1.24) 4.98 (3.74–6.63) 0.23 (0.14–0.36)
Control – 2.84 (2.28–3.54) 0.25 (0.14–0.44) 4.89 (4.10–5.83) 0.18 (0.15–0.22)
Janus NBSBWG (19) Case – – – 6.55 (6.10–7.03) –
Control – – – 5.82 (5.58–6.07) –
MDCS (24) Case – – – 9.57 (8.90–10.3) –
Control – – – 9.58 (9.06–10.1) –
NSHDC (25) Case – – – 15.5 (13.6–17.6) –
Control – – – 14.6 (13.4–15.9) –
Open in a new tab

Abbreviations: EPIC, European Prospective Investigation into Cancer and Nutrition; NBSBWG, Janus Nordic Biological Specimen Biobank Working Group; JACC, Japan Collaborative Cohort Study; JPHC, Japan Public Health Center‐based prospective Study; MDCS, Malmö Diet and Cancer Study; NSHDC, Northern Sweden Health and Disease Cohort.

The ORs for prostate cancer risk by fourths of circulating isoflavone concentrations (separately by Japanese and European studies), adjusted for age, marital status, educational attainment, smoking, height and BMI are shown in Figure 1. In men from Japan, there was some evidence of an inverse association between circulating equol and prostate cancer risk (multivariable‐adjusted OR for the highest versus lowest fourth = 0.61, 95% confidence interval [CI] = 0.39–0.97), although there was no significant trend (OR per 75 percentile increase = 0.69, 95 CI = 0.46–1.05, p trend = 0.085). Genistein and daidzein concentrations were not significantly associated with risk (ORs for the highest versus lowest fourth = 0.70, 0.45–1.10 and 0.71, 0.45–1.12, respectively). In men from Europe, circulating concentrations of genistein, daidzein and equol were not associated with risk.

Figure 1.

Figure 1

Open in a new tab

ORs for prostate cancer associated with isoflavone concentrations in Japanese (a) and European (b) studies. The black squares indicate the ORs in study‐specific fourths, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The diamonds show the OR for an increase in concentration from the 12.5 and 87.5 percentage points, and the widths of the diamonds show the 95% CIs. The χ2 tests for linear trend (Ptr) were calculated scoring the fourths as 0, 0.33, 0.67 and 1. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9 and ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180 and >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown) and cigarette smoking (never, past, current, unknown).

Figure 2 and Figures S1 and S2 show the associations of isoflavones with prostate cancer risk for each study. Note that because of the small number of cases in some studies and missing data on some covariates, these analyses were not adjusted for anthropometric and lifestyle factors; therefore, the results shown in these figures for all studies combined differ slightly from those shown in Figure 1. For genistein, there was evidence of heterogeneity within European studies (p heterogeneity = 0.001): the OR for a 75 percentile increase in genistein was 0.73 (95% CI: 0.56–0.95) in EPIC Phase 1, while it was 1.51 (95% CI: 1.06–2.13) in EPIC phase 2 (Fig. S1).

Figure 2.

Figure 2

Open in a new tab

Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in equol concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) × (linear trend) interaction term. Abbreviations: European Prospective Investigation into Cancer and Nutrition (EPIC), Japan Collaborative Cohort Study (JACC), Japan Public Health Center‐based prospective Study (JPHC).

The associations of isoflavone concentrations with prostate cancer risk subdivided by time to diagnosis and other characteristics are shown in Figures S3–S8. In Japanese studies, the associations of genistein and equol concentrations with prostate cancer risk did not differ significantly by age at diagnosis, time to diagnosis, year of diagnosis, age at blood collection, smoking status or alcohol consumption (Figs. S3 and S5). For daidzein, there was significant heterogeneity by year of diagnosis (p heterogeneity = 0.032) with a reduction in risk of prostate cancer for cases diagnosed before the year 2000 (OR per 75 percentile increase = 0.43, 95% CI = 0.20–0.92) but not in cases diagnosed later (OR per 75 percentile increase = 1.21, 95% CI = 0.69–2.12) (Fig. S4).

In European studies, we found no evidence of heterogeneity for the association of isoflavones with prostate cancer according to time to diagnosis or other characteristics (Figs. S6–S8), with the exception of daidzein and BMI (p heterogeneity = 0.028). Here, daidzein was associated with a reduced risk of prostate cancer in men with a BMI ≥ 25 kg/m2 (OR per 75 percentile increase = 0.72, 95% CI = 0.52–0.98) but not in normal‐weight men (OR per 75 percentile increase = 1.29, 95% CI = 0.84–2.00) (Fig. S7).

No associations between circulating lignan concentrations and prostate cancer risk were observed (OR per 75 percentile increase = 1.04, 95 CI = 0.91–1.19 for enterolactone and OR per 75 percentile increase = 0.99, 95 CI = 0.79–1.25 for enterodiol) (Fig. 3 ). There was also no evidence of heterogeneity in these associations between the different studies (Figs. S9 and S10) and these associations did not differ by tumour characteristics or lifestyle factors (Figs. S11 and S12).

Figure 3.

Figure 3

Open in a new tab

ORs for prostate cancer associated with lignan concentrations. The black squares indicate the ORs in study‐specific fourths, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The diamonds show the OR for an increase in concentration from the 12.5 and 87.5 percentage points, and the widths of the diamonds show the 95% CIs. The χ2 tests for linear trend (Ptr) were calculated scoring the fourths as 0, 0.33, 0.67 and 1. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9 and ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180 and >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting and unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown) and cigarette smoking (never, past, current and unknown).

4. Discussion

The results of this collaborative analysis of individual participant data show no strong evidence that prediagnostic circulating concentrations of isoflavones or lignans are associated with prostate cancer risk. Equol concentrations were weakly inversely related with overall prostate cancer risk in Japanese populations; this association was significant when we compared the highest with the lowest fourth, but we did not find a significant linear trend. When we looked into subtypes, we found some evidence of heterogeneity in three subgroups; however, given the number of comparisons made in our study, these may be chance findings.

This large collaborative analysis has brought together and reanalysed almost all of the available prospective data on the associations of circulating isoflavone and lignan concentrations with prostate cancer incidence, representing, as far as we are aware, all of the worldwide prospective data on circulating genistein, daidzein and equol, and over 90% of the worldwide data on circulating lignans. While data on enterolactone from the α‐Tocopherol, β‐Carotene Cancer Prevention Study (214 case patients)25 and the Helsinki Heart Study (136 case patients)18 were not available for this analysis, their results do not differ materially from those reported here, and it is unlikely that these data would have changed our summary risk estimates. The Multiethnic Cohort (MEC) study,27 in which phytoestrogen concentrations were measured in urine, found a weak inverse association between genistein and daidzein concentrations and prostate cancer risk, which did not differ across the four ethnic groups examined (African Americans, Japanese Americans, Latinos, Whites). However, this study was not included in these analyses because urine concentrations are not comparable to circulating concentrations, and because there was no other study with urinary phytoestrogen measures with which to pool these data.

Some studies have studies whether prostate cancer risk is related to the dietary intake of soya foods, but a meta‐analysis of prospective studies reported no association.28 A meta‐analysis of randomised controlled trials on the effect of soya supplementation (mainly combining genistein, daidzein and glycitein) on prostate cancer risk found only two small trials in which the endpoint included precursor lesions.29 Thus, conclusions on prostate cancer prevention cannot be drawn.

Equol is a metabolite produced in the gastrointestinal tract by bacterial metabolism of daidzein, and there is some evidence that equol has higher hormonal activity that genistein and daidzein.30, 31 In vitro and animal studies have shown that equol has anti‐androgenic and anti‐proliferative properties in the prostate,30, 32 but its role in humans remains unclear. Equol production varies between individuals, and the term equol‐producers refers to people who produce equol after consuming isoflavones; therefore it may be possible that the health benefits of soya‐based diets are greater in equol‐producers than in equol non‐producers.33. Only between 20 and 50% of individuals can produce equol,34, 35 and the capacity to produce equol has been found to be lower among American than Asian men,36 perhaps because Asian populations consume dietary isoflavones from an early age.

Both enterolactone and enterodiol are formed from plant‐lignan glycoside precursors by the activity of the gut microbiota in the proximal colon.31 Both in vitro and clinical studies have suggested that a diet rich in precursors of mammalian lignans (e.g., whole grain rye, bran or flaxseed) may have antiproliferative effects in the early stages of prostate cancer development.37, 38, 39 However, findings from the current study do not support the hypothesis that circulating lignans play a role in the aetiology of prostate cancer.

The current analyses have several strengths and limitations. Strengths include the large sample size, the detailed data on participant characteristics, and the consistent statistical approach used to analyse the individual participant data across the studies. Moreover, this collaboration only includes prospective studies, which allowed us to assess whether associations varied by time from blood collection to diagnosis and hence examine possible reverse causality. The limitations of these analyses include the use of a single blood sample measurement of phytoestrogens, as the true exposure of interest is medium to long‐term average levels of circulating phytoestrogens. Two studies in women have shown that the within‐person reproducibility (over a 1‐ to 3‐year period) is relatively poor for daidzein and genistein.40, 41 To our knowledge, only one such reproducibility study has been carried out in men, which showed an intra‐class correlation coefficient (ICC) for plasma genistein of 0.32 (95% CI 0.14–0.50) for samples collected approximately 5 years apart.16 However, these studies were performed in Western populations, where the consumption of isoflavones is usually low and episodic over time; therefore a single measurement of circulating isoflavones is unlikely to represent long‐term average exposure and may have led to attenuation of risk estimates. Low to moderate ICCs over several years (over a 1‐ to 3‐year period) have also been found for circulating enterolactone (ICCs of 0.5240 and 0.5541) and enterodiol (ICCs of 0.3741) in women. Although we have included almost all of the worldwide prospective data on circulating phytoestrogens and prostate cancer, we had limited power to assess associations by disease aggressiveness and other important sub‐groups, especially in Japanese populations, where isoflavone intake is high. Our study did not have sufficient mortality data to perform separate analyses using prostate cancer death as the outcome, but a recent cohort study found no association between prediagnostic enterolactone concentrations and mortality among men diagnosed with prostate cancer.42.

5. Conclusion

There was no strong evidence that prediagnostic circulating concentrations of isoflavones or lignans are associated with prostate cancer risk. However, further data are needed to examine these associations by disease aggressiveness, especially in populations following traditional East Asian diets.

Supporting information

Supplementary Table 1. Study characteristics, including sample population, recruitment and assessment characteristics.

Supplementary Table 2. Assay details

Supplementary Figure 1. Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in genistein concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) x (linear trend) interaction term. Abbreviations: European Prospective Investigation into Cancer and Nutrition (EPIC), Japan Collaborative Cohort Study (JACC), Japan Public Health Center‐based prospective Study (JPHC).

Supplementary Figure 2. Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in daidzein concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) x (linear trend) interaction term. Abbreviations: European Prospective Investigation into Cancer and Nutrition (EPIC), Japan Collaborative Cohort Study (JACC), Japan Public Health Center‐based prospective Study (JPHC).

Supplementary Figure 3. ORs for prostate cancer associated with genistein concentration, according to characteristics of cases and controls in Japanese studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of genistein concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of genistein with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 4. ORs for prostate cancer associated with daidzein concentration, according to characteristics of cases and controls in Japanese studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of daidzein concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of daidzein with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 5. ORs for prostate cancer associated with equol concentration, according to characteristics of cases and controls in Japanese studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of equol concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of equol with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 6. ORs for prostate cancer associated with genistein concentration, according to characteristics of cases and controls in European studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of genistein concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of genistein with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 7. ORs for prostate cancer associated with daidzein concentration, according to characteristics of cases and controls in European studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of daidzein concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of daidzein with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI = <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 8. ORs for prostate cancer associated with equol concentration, according to characteristics of cases and controls in European studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of equol concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of equol with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI = <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 9. Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in enterolactone concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) x (linear trend) interaction term. Abbreviations: European Prospective Investigation into Cancer and Nutrition (EPIC), Janus Nordic Biological Specimen Biobank Working Group (NBSBWG), the Malmö Diet and Cancer Study (MDCS), Northern Sweden Health and Disease Cohort (NSHDC)

Supplementary Figure 10. Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in enterodiol concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) x (linear trend) interaction term. Abbreviation: European Prospective Investigation into Cancer and Nutrition (EPIC).

Supplementary Figure 11. ORs for prostate cancer associated with enterolactone concentration, according to characteristics of cases and controls. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of enterolactone concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of enterolactone with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study, but not further adjusted.

Supplementary Figure 12. ORs for prostate cancer associated with enterodiol concentration, according to characteristics of cases and controls. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of enterodiol concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of enterodiol with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study, but not further adjusted.

Click here for additional data file. (2.3MB, docx)

Acknowledgments

The authors thank the men who participated in the collaborating studies, the research staff, collaborating laboratories and funding agencies in each of the studies.

Grant sponsor: Centralized pooling, checking and data analysis was supported by Cancer Research UK; Grant numbers: C8221/A19170 and C570/A11691. The funders of the study had no role in the study design, data collection, data analysis, data interpretation or writing of the report. The authors in the writing team had full access to all data in the study. The corresponding author had the final responsibility for the decision to submit for publication. The EHNBPCCG data are not suitable for sharing, but are suitable for research involving further analyses in collaboration with the Cancer Epidemiology Unit, Nuffield Department of Population Health, University of Oxford. For information on how to submit an application for gaining access to EPIC data and/or biospecimens, please follow the instructions at http://epic.iarc.fr/access/index.php. Details of funding for the original studies are in the relevant publications. Grant sponsors: EPIC: EPIC‐Greece: the Hellenic Health Foundation. EPIC‐Italy: Associazione Italiana per la Ricerca sul Cancro‐AIRC‐Italy. MORGEN‐EPIC cohort of the EPIC Bilthoven centre: Dutch Ministry of Public Health, Welfare and Sports (VWS), Netherlands Cancer Registry (NKR) and Statistics Netherlands (the Netherlands). EPIC‐Denmark: Danish Cancer Society. EPIC‐Norfolk: Medical Research Council and Cancer Research UK. Janus NBSBWG: The Norwegian Institute of Public Health for access to survey data in this study. The Nordic Cancer Union; Phytoprevent Project of the European Commission; Sigrid Juselius Foundation, Helsinki, Finland; Grant sponsor: JACC: Grant‐in‐Aid for Scientific Research on Priority Areas (2) from the Ministry of Education, Culture, Sports, Science and Technology of Japan; Grant number: 14031221; Grant sponsor: The JACC Study has also been supported by Grants‐in‐Aid for Scientific Research; Grant numbers: 61010076, 62010074, 63010074, 1010068, 2151065, 3151064, 4151063, 5151069, 6279102 and 11181101 from the same Ministry; Grant sponsor: JPHC: Grants‐in‐Aid for Cancer Research; Grant number: 19shi‐2; Grant sponsor: for the 3rd Term Comprehensive 10‐Year Strategy for Cancer Control; Grant number: H18‐sanjigan‐ippan‐001; Grant sponsor: for Research on Risk of Chemical Substances; Grant number: H17‐kagakuippan‐014) from the Ministry of Health, Labour and Welfare of Japan; Grant sponsor: Grants‐in‐Aid for Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports, Science and Technology; Grant number: 17015049; Grant sponsors: MDCS: Funding was obtained from the Swedish Council for Working Life and Social Research, Swedish Cancer Society, the Albert Påhlsson Foundation for Scientific Research, the Gunnar Nilsson Cancer Foundation, Skåne University Hospital – Foundations and Donations, the Malmö General Hospital Foundation for the Combating of Cancer and the Ernhold Lundström Foundation for Scientific Research. NSHDC: Swedish Cancer foundation, Project No. 4620, the European Union, Project No. QRLT 2000–00266.

Conflict of interest: The authors declare that they have no conflicts of interest.

References

  • 1. Pavese JM, Krishna SN, Bergan RC. Genistein inhibits human prostate cancer cell detachment, invasion, and metastasis. Am J Clin Nutr 2014;100(Suppl 1):431S–6S. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Yan L, Spitznagel EL. Soy consumption and prostate cancer risk in men: a revisit of a meta‐analysis. Am J Clin Nutr 2009;89:1155–63. [DOI] [PubMed] [Google Scholar]
  • 3. Mahmoud AM, Yang W, Bosland MC. Soy isoflavones and prostate cancer: a review of molecular mechanisms. J Steroid Biochem Mol Biol 2014;140:116–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Kurzer MS, Xu X. Dietary phytoestrogens. Annu Rev Nutr 1997;17:353–81. [DOI] [PubMed] [Google Scholar]
  • 5. Smeds AI, Eklund PC, Sjoholm RE, et al. Quantification of a broad spectrum of lignans in cereals, oilseeds, and nuts. J Agric Food Chem 2007;55:1337–46. [DOI] [PubMed] [Google Scholar]
  • 6. Landete JM. Plant and mammalian lignans: a review of source, intake, metabolism, intestinal bacteria and health. Food Res Int 2012;46:410–24. [Google Scholar]
  • 7. Larkin T, Price WE, Astheimer L. The key importance of soy isoflavone bioavailability to understanding health benefits. Crit Rev Food Sci Nutr 2008;48:538–52. [DOI] [PubMed] [Google Scholar]
  • 8. Selma MV, Espin JC, Tomas‐Barberan FA. Interaction between phenolics and gut microbiota: role in human health. J Agr Food Chem 2009;57:6485–501. [DOI] [PubMed] [Google Scholar]
  • 9. Endogenous Hormones Prostate Cancer Collaborative Group , Roddam AW, Allen NE, et al. Endogenous sex hormones and prostate cancer: a collaborative analysis of 18 prospective studies. J Natl Cancer Inst 2008;100:170–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Travis RC, Appleby PN, Martin RM, et al. A meta‐analysis of individual participant data reveals an association between circulating levels of IGF‐I and prostate cancer risk. Cancer Res 2016;76:2288–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Allen NE, Travis RC, Appleby PN, et al. Selenium and prostate cancer: analysis of individual participant data from fifteen prospective studies. J Natl Cancer Inst 2016;108:1–153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Crowe FL, Appleby PN, Travis RC, et al. Circulating fatty acids and prostate cancer risk: individual participant meta‐analysis of prospective studies. J Natl Cancer Inst 2014;106:1–240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Key TJ, Appleby PN, Travis RC, et al. Endogenous Hormones Nutritional Biomarkers Prostate Cancer Collaborative G. Carotenoids, retinol, tocopherols, and prostate cancer risk: pooled analysis of 15 studies. Am J Clin Nutr 2015;102:1142–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Price AJ, Travis RC, Appleby PN, et al. Circulating folate and vitamin B‐12 and risk of prostate cancer: a collaborative analysis of individual participant data from six cohorts including 6875 cases and 8104 controls. Eur Urol 2016;70:941–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Travis RC, Spencer EA, Allen NE, et al. Plasma phyto‐oestrogens and prostate cancer in the European Prospective Investigation into Cancer and Nutrition. Br J Cancer 2009;100:1817–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Travis RC, Allen NE, Appleby PN, et al. Prediagnostic concentrations of plasma genistein and prostate cancer risk in 1,605 men with prostate cancer and 1,697 matched control participants in EPIC. Cancer Cause Control 2012;23:1163–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Ward H, Chapelais G, Kuhnle GG, et al. Lack of prospective associations between plasma and urinary phytoestrogens and risk of prostate or colorectal cancer in the European Prospective into Cancer‐Norfolk study. Cancer Epidemiol Biomarkers Prev 2008;17:2891–4. [DOI] [PubMed] [Google Scholar]
  • 18. Stattin P, Adlercreutz H, Tenkanen L, et al. Circulating enterolactone and prostate cancer risk: a Nordic nested case‐control. Int J Cancer 2002;99:124–9. [DOI] [PubMed] [Google Scholar]
  • 19. Langseth H, Gislefoss RE, Martinsen JI, et al. Cohort profile: the Janus Serum Bank Cohort in Norway. Int J Epidemiol 2017;46:403–4g. [DOI] [PubMed] [Google Scholar]
  • 20. Hjerkind KV, Gislefoss RE, Tretli S, et al. Cohort profile update: the Janus Serum Bank Cohort in Norway. Int J Epidemiol 2017;46:1101–2f. [DOI] [PubMed] [Google Scholar]
  • 21. Ozasa K, Nakao M, Watanabe Y, et al. Serum phytoestrogens and prostate cancer risk in a nested case‐control study among Japanese men. Cancer Sci 2004;95:65–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Kurahashi N, Iwasaki M, Inoue M, et al. Plasma isoflavones and subsequent risk of prostate cancer in a nested case‐control study: the Japan Public Health Center. J Clin Oncol 2008;26:5923–9. [DOI] [PubMed] [Google Scholar]
  • 23. Wallstrom P, Drake I, Sonestedt E, et al. Plasma enterolactone and risk of prostate cancer in middle‐aged Swedish men. Eur J Nutr 2017. 10.1007/s00394-017-1530-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Stattin P, Bylund A, Biessy C, et al. Prospective study of plasma enterolactone and prostate cancer risk (Sweden). Cancer Cause Control 2004;15:1095–102. [DOI] [PubMed] [Google Scholar]
  • 25. Kilkkinen A, Virtamo J, Virtanen MJ, et al. Serum enterolactone concentration is not associated with prostate cancer risk in a nested case‐control study. Cancer Epidemiol Biomarkers Prev 2003;12:1209–12. [PubMed] [Google Scholar]
  • 26. Key TJ, Appleby PN, Allen NE, et al. Pooling biomarker data from different studies of disease risk, with a focus on endogenous hormones. Cancer Epidemiol Biomarkers Prev 2010;19:960–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Park SY, Wilkens LR, Franke AA, et al. Urinary phytoestrogen excretion and prostate cancer risk: a nested case‐control study in the Multiethnic Cohort. Br J Cancer 2009;101:185–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. WCRF/AICR . World Cancer Research Fund International/American Institute for Cancer Research Continuous Update Project Report: Diet, Nutrition, Physical Activity, and Prostate Cancer. Available at: http://www.wcrf.org/sites/default/files/Prostate-Cancer-SLR-2014.pdf 2014.
  • 29. van Die MD, Bone KM, Williams SG, et al. Soy and soy isoflavones in prostate cancer: a systematic review and meta‐analysis of randomized controlled trials. BJU Int 2014;113:E119–30. [DOI] [PubMed] [Google Scholar]
  • 30. Shor D, Sathyapalan T, Atkin SL, et al. Does equol production determine soy endocrine effects? Eur J Nutr 2012;51:389–98. [DOI] [PubMed] [Google Scholar]
  • 31. Adlercreutz H. Phyto‐oestrogens and cancer. Lancet Oncol 2002;3:364–73. [DOI] [PubMed] [Google Scholar]
  • 32. Lund TD, Munson DJ, Haldy ME, et al. Equol is a novel anti‐androgen that inhibits prostate growth and hormone feedback. Biol Reprod 2004;70:1188–95. [DOI] [PubMed] [Google Scholar]
  • 33. Setchell KD, Clerici C. Equol: history, chemistry, and formation. J Nutr 2010;140:1355S–62S. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Rowland IR, Wiseman H, Sanders TA, et al. Interindividual variation in metabolism of soy isoflavones and lignans: influence of habitual diet on equol production by the gut microflora. Nutr Cancer 2000;36:27–32. [DOI] [PubMed] [Google Scholar]
  • 35. Arai Y, Uehara M, Sato Y, et al. Comparison of isoflavones among dietary intake, plasma concentration and urinary excretion for accurate estimation of phytoestrogen intake. J Epidemiol 2000;10:127–35. [DOI] [PubMed] [Google Scholar]
  • 36. Akaza H, Miyanaga N, Takashima N, et al. Comparisons of percent equol producers between prostate cancer patients and controls: case‐controlled studies of isoflavones in Japanese, Korean and American residents. Jpn J Clin Oncol 2004;34:86–9. [DOI] [PubMed] [Google Scholar]
  • 37. Landberg R, Andersson SO, Zhang JX, et al. Rye whole grain and bran intake compared with refined wheat decreases urinary C‐peptide, plasma insulin, and prostate specific antigen in men with prostate cancer. J Nutr 2010;140:2180–6. [DOI] [PubMed] [Google Scholar]
  • 38. McCann MJ, Rowland IR, Roy NC. The anti‐proliferative effects of enterolactone in prostate cancer cells: evidence for the role of DNA licencing genes, mi‐R106b cluster expression, and PTEN dosage. Nutrients 2014;6:4839–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. McCann MJ, Gill CI, McGlynn H, et al. Role of mammalian lignans in the prevention and treatment of prostate cancer. Nutr Cancer 2005;52:1–14. [DOI] [PubMed] [Google Scholar]
  • 40. Kotsopoulos J, Tworoger SS, Campos H, et al. Reproducibility of plasma and urine biomarkers among premenopausal and postmenopausal women from the nurses' health studies. Cancer Epidem Biomar 2010;19:938–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Zeleniuch‐Jacquotte A, Adlercreutz H, Akhmedkhanov A, et al. Reliability of serum measurements of lignans and isoflavonoid phytoestrogens over a two‐year period. Cancer Epidem Biomar 1998;7:885–9. [PubMed] [Google Scholar]
  • 42. Eriksen AK, Kyro C, Norskov N, et al. Prediagnostic enterolactone concentrations and mortality among Danish men diagnosed with prostate cancer. Eur J Clin Nutr 2017;71:1235–1240. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Table 1. Study characteristics, including sample population, recruitment and assessment characteristics.

Supplementary Table 2. Assay details

Supplementary Figure 1. Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in genistein concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) x (linear trend) interaction term. Abbreviations: European Prospective Investigation into Cancer and Nutrition (EPIC), Japan Collaborative Cohort Study (JACC), Japan Public Health Center‐based prospective Study (JPHC).

Supplementary Figure 2. Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in daidzein concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) x (linear trend) interaction term. Abbreviations: European Prospective Investigation into Cancer and Nutrition (EPIC), Japan Collaborative Cohort Study (JACC), Japan Public Health Center‐based prospective Study (JPHC).

Supplementary Figure 3. ORs for prostate cancer associated with genistein concentration, according to characteristics of cases and controls in Japanese studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of genistein concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of genistein with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 4. ORs for prostate cancer associated with daidzein concentration, according to characteristics of cases and controls in Japanese studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of daidzein concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of daidzein with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 5. ORs for prostate cancer associated with equol concentration, according to characteristics of cases and controls in Japanese studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of equol concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of equol with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 6. ORs for prostate cancer associated with genistein concentration, according to characteristics of cases and controls in European studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of genistein concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of genistein with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI; <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 7. ORs for prostate cancer associated with daidzein concentration, according to characteristics of cases and controls in European studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of daidzein concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of daidzein with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI = <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 8. ORs for prostate cancer associated with equol concentration, according to characteristics of cases and controls in European studies. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of equol concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of equol with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study and adjusted for age at blood collection (exact), body mass index (BMI = <25, 25–27.4, 27.5–29.9, ≥30 kg/m2, unknown), height (≤170, 171–175, 176–180, >180 cm, unknown), marital status (married/cohabiting, not married/cohabiting, unknown), educational status (did not graduate from high school/secondary school/college, high school/secondary school/college graduates, university graduates, unknown), and cigarette smoking (never, past, current, unknown).

Supplementary Figure 9. Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in enterolactone concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) x (linear trend) interaction term. Abbreviations: European Prospective Investigation into Cancer and Nutrition (EPIC), Janus Nordic Biological Specimen Biobank Working Group (NBSBWG), the Malmö Diet and Cancer Study (MDCS), Northern Sweden Health and Disease Cohort (NSHDC)

Supplementary Figure 10. Study‐specific ORs (95% CIs) for prostate cancer associated with a 75 percentile increase in enterodiol concentrations. Estimates are from logistic regression conditioned on the matching variables within each study, but not further adjusted. Heterogeneity in linear trends between studies and between Japanese and European studies was tested by comparing the χ2 values for models with and without a (studies) x (linear trend) interaction term. Abbreviation: European Prospective Investigation into Cancer and Nutrition (EPIC).

Supplementary Figure 11. ORs for prostate cancer associated with enterolactone concentration, according to characteristics of cases and controls. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of enterolactone concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of enterolactone with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study, but not further adjusted.

Supplementary Figure 12. ORs for prostate cancer associated with enterodiol concentration, according to characteristics of cases and controls. Each OR is the estimate of the linear trend obtained by replacing the categorical variables representing the fourths of enterodiol concentration by a continuous variable scored as 0, 0.33, 0.67, and 1. Black squares indicate the OR, and the horizontal lines show the 95% CIs. The area of each square is proportional to the amount of statistical information (inverse of the variance of the logarithm of the OR). The vertical dotted line indicates the OR for all studies. Tests for heterogeneity are for the difference in the association of enterodiol with prostate cancer risk between subgroups. Estimates are from conditional logistic regression on case‐control sets matched within each study, but not further adjusted.

Click here for additional data file. (2.3MB, docx)

Articles from International Journal of Cancer are provided here courtesy of Wiley

RESOURCES