Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.

EPI-10-0752
Cancer
Epidemiology,
Biomarkers
Research Article & Prevention

Dietary Phytoestrogens and the Risk of Ovarian Cancer in

the Women's Lifestyle and Health Cohort Study
€ f3,4, Therese M.-L. Andersson3, Herman Adlercreutz5, and Elisabete Weiderpass1,3,6
Maria Hedelin1,2, Marie Lo

Abstract
Background: Dietary intake of phytoestrogens has been inversely associated to hormone-dependent
cancers, such as prostate and breast cancers. Few studies have investigated the association between ovarian
cancer and intake of phytoestrogens. We evaluated the associations between intake of phytoestrogens
(isoflavonoids/lignans/coumestrol) and fiber (vegetable/cereal) and risk of ovarian cancer.
Methods: In 1991–1992 a prospective population-based cohort study among Swedish women was con-
ducted, including 47,140 women with complete dietary questionnaire data. During follow-up until Decem-
ber 2007, 163 women developed invasive (n ¼ 117) and borderline (n ¼ 46) ovarian cancers. The median
follow-up time was 16 years and total person year was 747,178. Cox proportional hazards models were
conducted to estimate multivariate risk ratios, 95% CI for associations with risk of ovarian cancer.
Results: We found no association between intake of phytoestrogens or fiber and overall ovarian cancer risk.
In addition, we found no statistically significant association between intake of specific food items rich in
phytoestrogens (berries, nuts, beans/soy, and crisp or whole-grain bread) and ovarian cancer risk overall. Fiber
and coumestrol was inversely associated with borderline ovarian cancer, but not with invasive ovarian cancer.
Conclusions: We found no association between intake of phytoestrogens or fiber and overall ovarian
cancer risk.
Impact: Phytoestrogens do not play a major etiologic role in ovarian cancer, at least among women in this
Swedish cohort with low bean/soy intake. However, our results of a difference in the effect of fiber or
coumestrol between invasive and borderline ovarian cancer need to be evaluated in larger studies.
Cancer Epidemiol Biomarkers Prev; 20(2); 308–17.
2011 AACR.

Introduction Isoflavonoids, a class of phytoestrogens found in soy-

based foods, have estrogenic, antiestrogenic, and anti-
Phytoestrogens, compounds naturally found in plant proliferative effects and inhibit the growth and prolifera-
foods and structurally related to endogenous estrogens, tion of ovarian cancer (5–8). Results from a meta-analysis
have been inversely associated to hormone-dependent showed that high intake of isoflavonoids or soy-based
cancers, such as prostate and breast cancers (1, 2). Few food were associated with a decreased risk of ovarian
have investigated the relationship between ovarian can- cancer (4). Half of these studies were conducted in an
cer and intake of phytoestrogens (3, 4). Asian population were the intake of soy products are
much higher than in a Western population. Dietary
sources of isoflavonoids in Western populations are
based on intake of other beans, pea, and various other
Authors' Affiliations: 1Department of Genetic Research, Folkha €lsan vegetables and fruits. However, there is no clear etiologic
Research Center, Biomedicum Helsinki, Helsinki, Finland; 2Unit of Clinical role of vegetable intake in ovarian cancer (9). Many
Cancer Epidemiology, Division of Oncology, Department of Clinical
Sciences, The Sahlgrenska Academy, University of Gothenburg, Gothen-
studies showed no association between vegetable intake
burg, Sweden; 3Department of Medical Epidemiology and Biostatistics, and ovarian cancer risk (10–12), some found a protective
Karolinska Institutet, Stockholm, Sweden; 4Department of Clinical and effect (3, 13, 14), whereas some found an increased risk
Experimental Medicine, University of Linko€ ping, Linko
€ ping, Sweden; 5Pro-
gram for Preventive Medicine, Folkha€lsan Research Center, and Division of with high vegetable intake (15).
Clinical Chemistry, University of Helsinki, Helsinki, Finland; and 6Depart- Lignans, another class of phytoestrogens, have a lower
ment of Etiological Research, Cancer Registry of Norway, Oslo and Depart- estrogenic activity than isoflavonoids (16), but have also
ment of Community Medicine, University of Tromsø, Tromsø, Norway
antiestrogenic and anticarcinogenic effects (17). Com-
Corresponding Author: Elisabete Weiderpass, Department of Medical
Epidemiology and Biostatistics, Karolinska Institutet, PO Box 281, Stock- pared with isoflavonoids, the intake of lignans is more
holm 17177, Sweden. Phone: 840-845-3406. E-mail: in Western populations; some of the human dietary
Elisabete.Weiderpass.Vainio@ki.se sources of lignans are flaxseed, grain (especially rye),
doi: 10.1158/1055-9965.EPI-10-0752 seeds, and berries (18). In one small American case–control

2011 American Association for Cancer Research. study, high intake of the lignan compound, matairesinol

308 Cancer Epidemiol Biomarkers Prev; 20(2) February 2011

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Dietary Phytoestrogens and the Risk of Ovarian Cancer

(MAT) and secoisolariciresinol (SECO), were found to be into coumestans, isoflavonoids, and lignans. Coumestrol
inversely related to ovarian cancer risk (3). (a coumestan) and the isoflavonoids genistein, daidzein,
To our knowledge, only 1 cohort study (11) and 2 case– and their plant precursors biochanin A and formonone-
control (3, 19) studies have examine the intake of phy- tin, are mainly found in soybeans and clover. Plant
toestrogens in relation to ovarian cancer in a Western lignans, such as MAT and SECO, are converted by the
setting and no one have investigated the phytoestrogen colonic microflora to mammalian lignans, now entero-
compound, coumestrol. Furthermore, no study has yet lignans (23), enterolactone, and enterodiol (20). Until
included the newly identified lignan precursors laricir- recently, only 2 plant lignan precursors for mammalian
esinol, pinoresinol, syringaresinol, and medioresinol lignans were known: SECO and MAT. Lariciresinol,
when calculating total intake of lignans. Lignans are pinoresinol, syringaresinol, and medioresinol are newly
especially found in the fiber-rich outer layer of cereal identified enterolactone precursors, found mainly in cer-
grains (20), and a diet high in fiber interferes with the eals and seeds.
enterohepatic metabolism of estrogens reducing their
levels in the body (21). Use of antibiotics could affect Exposure assessment
the serum levels of bioactive phytoestrogens, because the Known and potential risk factors for ovarian cancer
colonic microflora is involved in the biotransformation of were assessed with a self-administered questionnaire,
phytoestrogens (22, 23). including average intake of foods and beverages (24).
In this prospective, population-based study of Swedish Dietary habits during the 6 months before the enrolment
women we examined the association between risk of in the study were ascertained through a validated semi-
ovarian cancer and dietary intake of total isoflavonoids, quantitative food-frequency questionnaire (FFQ) that
lignans, and coumestrol. We also evaluated the associa- covered the frequency of consumption and quantity of
tion between dietary intake of total fiber, cereal, and approximately 80 food items and beverages, including 50
vegetable fiber and the risk of ovarian cancer. items containing phytoestrogens (26). For example, par-
ticipants were asked how often, on average, they ate bean,
Subjects and Methods soy, or lentil; green pea; pea soup; broccoli; cauliflower;
white or red cabbage; spinach; onion or leek; carrot;
Study population swede or beetroot; porridge or gruel; wheat or oat bran;
The study design and exposure assessment have been cereals or m€ usli, nut or almond; apple or pear: never/
previously described in detail (24, 25). The source popula- seldom, 1–3 times/month, 1 time/week, 2 times/week,
tion of this study was women ages 30 to 49 years, residing 3–4 times/week, 5–6 times/week, 1 time/day, 2 times/
in the Uppsala Health Care Region in Sweden between day, or 3 times/day. The participants were also asked
1991 and 1992. From this source population, 96,000 how many slices of bread they ate per day or week; wheat
women were randomly selected from 4 age strata (30– bread, whole meal bread (bread baked of coarse and
34, 35–39, 40–44, and 45–49 years) and were invited to whole meal flour), or crisp bread (mostly baked of rye
participate in the Swedish component of the Scandinavian meal). To estimate individual intake of energy and nutri-
Women’s Lifestyle and Health Cohort (WLH study). The ents, we linked the dietary information from the ques-
women were asked to fill in a questionnaire, and of those tionnaire to the nutrient database created by the Swedish
invited, 49,259 returned the questionnaires and were National Food Administration (food table of 1989; ref. 27).
enrolled in the study. The Swedish Data Inspection Board Furthermore, to estimate the intake of specific phytoes-
and the regional Ethical Committee approved the study. trogens, we created a database (25, 28) with information
Among the 49,248 Swedish women with dietary data from published analytical data for the content of isofla-
included in the study, 171 epithelial invasive (n ¼ 122) or vonoids (genistein, daidzein, biochanin A, formononetin,
borderline (n ¼ 49) ovarian cancer cases were diagnosed equol), coumestans (coumestrol), and lignans (MAT,
by the end of follow-up in December 2007. The following SECO) in food products. The content of the lignans
were excluded, in order: subject who have undergone lariciresinol, pinoresinol, syringaresinol, and medioresi-
bilateral oophorectomy (information from the baseline nol in different grain flours was used to estimate lignan
questionnaire) and therefore not at risk for ovarian cancer content of bread and cereal products, and added to the
(n ¼ 433), those with energy intakes outside the first database. Information from recently published analytical
(1,852 kJ/day) and 99th (12,465 kJ/day) percentiles data for the content of lariciresinol and pinoresinol in
(n ¼ 977), and those with missing values in any of the vegetables and fruits were added to the database (17, 29–
adjustment covariates (n ¼ 696). A total of 47,140 women 31). Content of syringaresinol and medioresinol in vege-
(117 invasive and 46 borderline ovarian cancer cases) tables and fruits were available only for tomato, orange,
were included in the final analysis. The median follow- citrus, strawberry, and lingonberry (17). On the basis of
up time was 16 years and total person year was 747,178. the levels of genistein, daidzein, equol, MAT, enterodiol,
and enterolactone in raw cow milk, we estimated the
Classification of phytoestrogens content of these compounds in different kinds of milk
Phytoestrogens are naturally occurring hormone-like products, and the information was added to the database
compounds found in plant food and can be subdivided (28). Most of the analyses of phytoestrogen compounds in

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 20(2) February 2011 309

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Hedelin et al.

food products were carried out by isotope dilution gas Food items with the highest lignan, isoflavonoid, or
chromatography–mass spectrometry performed in a coumestrol content per edible portion within the FFQ
laboratory in Finland (20, 31). Analyses of the content of were analyzed separately, and these food items were
lariciresinol and pinoresinol in vegetables and fruits were berries, nuts, beans (beans, soy, and lentil), or crisp
carried out by alkaline hydrolysis (29) followed by gas and whole-grain bread. All exposures except coumestrol,
chromatography–mass spectrometry (20, 31) or by liquid bean, and nuts were categorized into quartiles and for
chromatography–tandem mass spectrometry (29). In addi- each comparison of dietary intake; the lowest quartile
tion to the FFQ part of the questionnaire, participants were was used as the reference category. Coumestrol, beans,
asked how often, on average, they ate berries: times/week, and nuts were split into 3 categories because 50%–60% of
times/month, or seldom/never. The reported frequency the women had zero intakes, and they form the reference
together with an estimated portion size (115 g berries/ group. The remaining women were split into 2 equally
portion for all women) was used to calculate intake of large groups (cutoff at median: 0.014 mg/dayMJ for
berries (g/day). coumestrol, 0.73 g/dayMJ for beans, and 0.063 g/dayMJ
In 2002–2003, a second questionnaire was sent to those for nuts). However, the correlation between coumestrol
women who were alive and living in Sweden and of those and bean intake was 1.0 and the number of women were
34,415 answered the follow-up questionnaire. The parti- equally divided into the categories of the coumestrol and
cipants were asked about their use of antimicrobial med- the bean variable, indicating that the dietary source of
ication, "How many times in your life have you been coumestrol was exclusive beans, soy and lentils. There-
treated with penicillin or other antibiotics?" (never, 1–2 fore, only results for coumestrol are presented.
times, 3–10 times, more than 10 times). Age- and energy-adjusted models were fitted, and
models adjusted for additional potential confounders
Follow-up including body mass index (BMI), level of education,
Follow-up of the cohort was achieved through linkages country of birth, smoking, use of oral contraceptives,
with existing nationwide health registers in Sweden. age at menarche, parity, use of postmenopausal hormone
Because each resident in Sweden has a unique national replacement therapy, and selected food groups and nutri-
registration number, one can link the data from the cohort ent densities categorized into quartiles. Variables in the
with these registers for virtually complete follow-up with models were continues or categorized as defined in
respect to death and emigration. From the total popula- Table 1. We initially tested all covariates, and selection
tion registers, we received information on the dates of of the covariates to be included in the final models was
death for women who died during the follow-up period based on statistical significance and previous subject
and dates of emigration until December 31, 2007. The matter knowledge. Those included in the final models
national cancer registry provided data on prevalent can- were known risk factors or considered to be important
cer cases at cohort enrolment and on epithelial invasive or confounding factors for the relation between the main
borderline ovarian ((ICD7 ¼ 175) cancer diagnosed in the exposure and ovarian cancer, and are listed in the table
cohort during follow-up and SNOMED (Systematized footnotes (Tables 2 and 3). In analysis of the different
Nomenclature of Medicine) codes (Borderline: 84423, phytoestrogen compounds we adjusted mutually for
84513, 84623, 84723; invasive: 80203, 81403, 82603, other classes of phytoestrogen (isoflavonoids/lignans/
83103, 83803, 84403, 84413, 84603, 84703, 826031, coumestrol). Also, in analysis of different types of fiber
838031, 838032, 838033, 844131, 844133, 846032, 846033, we mutually adjusted cereal fiber or vegetable fiber;
847031, 847032). The start of follow-up was defined as the however, the risk estimates did not change and therefore
date of return of the questionnaire. Observation time was the variables were not included in the final model.
calculated from date of entry into the cohort until the As we have only incomplete information about meno-
occurrence of epithelial invasive or borderline ovarian pausal status after the start of follow-up, we considered
cancer, emigration, death, or end of the observation the average age at menopause in Sweden, 50 years (34), as
period (December 31, 2007). applicable for all cohort members. Menopausal status was
evaluated by fitting separate models for ovarian cancer
Statistical methods occurring at less or more than 50 years of age. Of the total
The association between phytoestrogens and risk of 163 ovarian cancer cases, 58 occurred before 50 years of
ovarian cancer was evaluated by Cox proportional age. When analyzing the time course starting at 50 years of
hazards models utilizing attained age as time scale age 36,634 women were available for analysis (when
(32). We interpreted relative hazards as estimates of censoring women with ovarian cancer before 50 years
relative risks (RR), given with 95% confidence intervals of age, women who died or emigrated or who were too
(CI). This corresponds to a 2-sided 5% level of signifi- young to reach 50 years at the end of follow-up). Among
cance. these 36,634 women, 105 ovarian cancer cases occurred.
Nutrient density was obtained through dividing the In addition, we tested for interactions between phy-
estimated intake of phytoestrogens (mg/day) and other toestrogens and use of antibiotics (2 times respective
nutrients or specific food items by the total energy intake >2 times in life) to evaluate whether the effect of
(MJ/day; Multivariate Nutrient Density Model; ref. 33). phytoestrogens was modified by antibiotic treatment.

310 Cancer Epidemiol Biomarkers Prev; 20(2) February 2011 Cancer Epidemiology, Biomarkers & Prevention

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Dietary Phytoestrogens and the Risk of Ovarian Cancer

Table 1. Selected baseline characteristics of participants with questionnaire data in 1991–1992 in the
Swedish Women's Lifestyle and Health Cohort Study

Cohort without Cohort with ovarian

ovarian cancer cancer
Characteristics (n ¼ 46,977) (n ¼ 163)

Age at entry, years, mean (SD) 40 (5.7) 43 (4.8)

BMI, kg/m2, mean (SD) 23.5 (3.7) 24 (3.4)
BMI, kg/m2, n (%)
<25 (normal weight) 33,919 (72) 101 (62)
25–29.9 (overweight) 8,875 (19) 45 (28)
30 (obese) 2,475 (5) 10 (6)
Missing 1,708 (4) 7 (4)
Years of education, n (%)
0–10 13,633 (29) 59 (36)
11–13 18,105 (38) 58 (36)
>13 14,351 (31) 44 (27)
Missing 888 (2) 2 (1)
Smoking status, n (%)
Never 19,004 (41) 62 (38)
Ever 27,830 (59) 101 (62)
Missing 143 (0.3) 0 (0)
Use of oral contraceptives, n (%)
Never 7,611 (16) 39 (24)
Ever 39,366 (84) 124 (76)
Parity, n (%)
Nulliparous 6,516 (14) 25 (15)
1 child 7,258 (15) 22 (14)
2 children 20,304 (43) 64 (39)
3 children 9,752 (21) 41 (25)
4 children 3,147 (7) 11 (7)
Use of antibiotics, lifetime, n (%)
Never 716 (2) 3 (2)
1–2 times 6,199 (13) 19 (12)
3–10 times 16,840 (36) 56 (34)
>10 times 9,078 (19) 20 (12)
Missing 14,144 (30) 65 (40)

Age at menarche, years, mean (SD) 13 (1.4) 13 (1.3)

Use of hormone replacement therapy
Never 45,499 (97) 154 (95)
Ever 1,478 (3) 9 (5)
Ovarian cancer subtypesa
Invasive - 117
Borderline - 46

Total energy intake, kJ, mean (SD) 6,520 (1,890) 6,140 (1,920)
Proportion of total energy intake,%
Fat 31 31
Protein 17 17
Carbohydrate 52 52
Alcohol 1 1
Dietary intake, g/day, mean (5th–95th percentiles)
Meat 80 (21–148) 74 (22–134)
Fish 19 (0–42) 17 (3–38)

(Continued on the following page)

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 20(2) February 2011 311

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Hedelin et al.

Table 1. Selected baseline characteristics of participants with questionnaire data in 1991–1992 in the
Swedish Women's Lifestyle and Health Cohort Study (Cont'd)

Cohort without Cohort with ovarian

ovarian cancer cancer
Characteristics (n ¼ 46,977) (n ¼ 163)

Vegetables 87 (20–181) 86 (13–185)

Fruit 124 (11–310) 121 (11–334)
Beans 3 (0–10) 3 (0–10)
Total fiber 14 (7–23) 14 (6–24)
Vegetable fiber 5 (2–11) 5 (2–11)
Cereal fiber 8 (3–14) 8 (3–15)
Dietary phytoestrogen intake, mg/day, mean (5th–95th percentiles)
Total isoflavonoidsb 74 (2–247) 69 (2–247)
Total lignansc 2,336 (1,019–4,007) 2,317 (1,054–4,356)
Coumestrol 0.05 (0–0.2) 0.05 (0–0.2)
a
See Subjects and Methods for definitions.
b
Including genistein, daidzein, formononetin, biochanin A, and equol.
c
Secoisolariciresinol (sum of anhydrosecoisolariciresinol and secoisolariciresinol), matairesinol, lariciresinol, isolariciresinol, pinor-
esinol, syringaresinol, and medioresinol.

Separate models were fitted for borderline respective total lignan, total isoflavonoid compounds, or coumestrol
invasive ovarian cancer. All analyses were done using (Table 2). There was no statistically significant association
the STATA version 10.1. with risk of ovarian cancer for specific food items rich in
phytoestrogens (berries, nuts, beans/soy, crisp and
Results whole-grain bread). For example, the RRs comparing
the highest to the lowest quartile of berry or crisp and
Characteristics of study participants whole-grain bread was 1.5 (95% CI: 0.95–2.32) and 1.48
Baseline characteristics and intake of nutrients among (95% CI: 0.95–2.31), respectively. The RRs comparing the
the study participants are presented in Table 1. As highest to the lowest category of nut intake was 0.88 (95%
expected, women who developed ovarian cancer tended CI: 0.56–1.38; data not shown).
to be older, use oral contraceptives less frequently and use We found no statistically significant association
hormone replacement therapy more often than women between intake of total dietary fiber, vegetable fiber, or
who did not develop ovarian cancer. Ninety-seven per- cereal fiber and ovarian cancer, although there was mar-
cent of the women were born in one of the Northern ginally significant decreased risk among women with an
countries (92% were born in Sweden; data not shown). intermediate intake of total fiber. For example, the RR
Intake of main groups of macronutrients and dietary comparing the third to the lowest quartile of fiber intake
mean intake of fiber, meat, fish, vegetables, fruit, or beans was 0.67 (95% CI: 0.42–1.08; ref. Table 2). In additional
were very similar between the 2 groups. The intake of analysis we found no association between risk of overall
phytoestrogens among women who developed ovarian ovarian cancer and intake of total carbohydrates, total
cancer did not differ substantially from the intake among cereals products, or total vegetables (data not shown).
women who did not develop ovarian cancer (Table 1). In When borderline and invasive ovarian cancer cases were
both groups, the mean daily intake of lignans was more examined separately, the RRs comparing the highest to the
than that of isoflavonoids and coumestrol and the highest lowest quartile of isoflavonoids, coumestrol, total fiber,
median intakes of lignans were seen for syringaresinol cereal fiber, or vegetable fiber intake were different
and medioresinol (data not shown). Rye bread, wheat between the 2 subtypes of ovarian cancer, with a tendency
bread, cereals, and berries contributed the most to the of a protective effect for borderline ovarian cancer but not
intake of lignans, whereas beans (bean, soy, and lentil) for invasive cancer (Table 3). For example, for intake of
were the most important dietary source of isoflavonoids coumestrol, there was a decreased risk for borderline
and exclusive source for coumestrol (data not shown). ovarian cancer among women with the highest intake of
coumestrol (RR ¼ 0.29; 95% CI: 0.09–0.94) compared with
Dietary phytoestrogens, fiber, and epithelial women with no intake; however, after multivariate adjust-
invasive or borderline ovarian cancer risk ment the inverse relationship was only marginally signifi-
We found no statistically significant association cant (RR ¼ 0.32; 95% CI: 0.10–1.07). The corresponding RR
between risk of ovarian cancer and dietary intakes of for invasive ovarian cancer was 1.47 (95% CI: 0.94–2.29),

312 Cancer Epidemiol Biomarkers Prev; 20(2) February 2011 Cancer Epidemiology, Biomarkers & Prevention

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Dietary Phytoestrogens and the Risk of Ovarian Cancer

Table 2. Risk of epithelial ovarian cancer in relation to estimated dietary intake of total lignans, total
isoflavonoids, coumestrol, and fiber, estimated as RR with 95% CI

Age and energy Multivariatec

adjusted

Mean (interquartile Cases/person years RRb 95% CIb RRb 95% CIb
Dietary intakea range)

Total lignans, 225 (13–282) 32/185,592 1.00 Reference 1.00 Reference

mg/dMJd
317 (283–351) 48/187,176 1.35 0.86–2.11 1.35 0.85–2.13
389 (352–431) 37/187,507 0.97 0.60–1.56 0.97 0.59–1.59
528 (432–1,713) 46/186,904 1.09 0.69–1.72 1.04 0.63–1.72
Total isoflavonoids, 0.5 (0–0.73) 34/186,437 1.00 Reference 1.00 Reference
mg/dMJe
0.9 (0.74–1.1) 38/187,165 1.07 0.67–1.70 1.09 0.67–1.77
6.5 (1.2–15) 48/186,852 1.37 0.88–2.13 1.47 0.90–2.38
38 (16–750) 43/187,725 1.15 0.74–1.81 1.26 0.79–2.01
Coumestrol, None 99/469,949 1.00 Reference 1.00 Reference
mg/dMJf
<0.014 32/139,436 1.20 0.79–1.81 1.25 0.82–1.90
0.014 32/137,794 1.03 0.69–1.54 1.10 0.73–1.65
Total fiber, g/dMJ 1.6 (0–1.82) 43/185,414 1.00 Reference 1.00 Reference
2.0 (1.83–2.1) 40/187,271 0.86 0.56–1.33 0.87 0.56–1.35
2.2 (2.2–2.4) 34/187,568 0.68 0.43–1.07 0.67 0.42–1.08
3.0 (2.5–10) 46/186,926 0.84 0.55–1.27 0.82 0.50–1.35
Cereal fiber, g/dMJ 0.8 (0–0.9) 35/186,112 1.00 Reference 1.00 Reference
1.1(1.0–1.2) 49/186,810 1.41 0.91–2.18 1.42 0.92–2.21
1.3 (1.3–1.4) 33/187,351 0.92 0.57–1.48 0.93 0.57–1.51
1.8 (1.5–5.7) 46/186,905 1.19 0.77–1.85 1.17 0.74–1.87
Vegetable fiber, g/dMJ 0.4 (0–0.57) 38/186,324 1.00 Reference 1.00 Reference
0.7 (0.58–0.78) 39/186,987 0.96 0.62–1.51 1.01 0.64–1.59
0.9 (0.79–1.0) 41/187,278 0.94 0.61–1.47 1.00 0.63–1.58
1.4 (1.1–9.5) 45/186,589 0.94 0.60–1.45 1.02 0.63–1.64
a
Exposure is categorized into quartiles and the lowest quartile was used as the reference category.
b
RR (95% CI) was obtained by Cox proportional hazards models.
c
Adjusted for age, oral contraceptives, age at menarche, parity, hormone replacement therapy, and intake of total energy intake,
alcohol, saturated fat, meat, and fish.
d
Secoisolariciresinol, matairesinol, lariciresinol, isolariciresinol, pinoresinol, syringaresinol, and medioresinol.
e
Including genistein, daidzein, formononetin, biochanin A, and equol.
f
Intake of coumestrol is categorized into 3 groups, where the reference group contains those with zero intakes.

and the point estimate for borderline ovarian cancer was food items rich in phytoestrogens, between the groups of
outside this CI, indicating a difference of the coumestrol antibiotic intake there was no significant heterogeneity
effect for the 2 subtypes of ovarian cancer. For lignans the for any association with phytoestrogen intake between
estimates were similar across sub-types of ovarian cancer the 2 groups of antibiotic intake. For example, the RR
(Table 3) and also for nuts, berries and crisp and whole- comparing the highest to the lowest quartile of crisp and
grain breads (data not shown). whole-grain bread intake was 0.4 (95% CI: 0.16–1.20) and
For a subset of the women with information on use of 1.3 (95% CI: 0.75–2.36) for those with low and high intake
antibiotic we carried out stratified analysis and examined of antibiotics, respectively (P ¼ 0.12; data not shown).
heterogeneity of the association with phytoestrogens and We repeated all analyses separately for women ages
ovarian cancer, between those with no or low intake of less or more than 50 years, and the estimates were similar
antibiotics (2 times in life) and those with high intake of across age groups (data not shown). However, for all
antibiotics (>2 times in life). Although the point estimates stratified analyses the sample sizes were small, leading to
were different for isoflavonoids, coumestrol, and specific unstable estimates.

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 20(2) February 2011 313

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 314
Table 3. Risk of epithelial invasive or borderline ovarian cancer in relation to estimated dietary intake of total lignans, total isoflavonoids,
coumestrol, and fiber, estimated as RR with 95% CI*. Exposures are categorized into quartiles and the lowest quartile was used as the
Hedelin et al.

reference categorya

Dietary intake Invasive ovarian cancer Borderline ovarian cancer

(mg/dayMJ)

Cases/ Age and energy Multivariatec Cases/ Age and energy Multivariatec
person adjusted person adjusted
years years
RRb 95% CIb RRb 95% CIb RRb 95% CIb RRb 95% CIb

Total lignansd 23/185,509 1.00 Reference 1.00 Reference 9/185,387 1.00 Reference 1.00 Reference
34/187,073 1.31 0.77–2.23 1.29 0.75–2.22 14/186,866 1.46 0.63–3.38 1.49 0.63–3.51
24/187,383 0.87 0.49–1.55 0.84 0.46–1.52 13/187,278 1.24 0.53–2.90 1.33 0.54–3.27
36/186,813 1.22 0.72–2.07 1.06 0.59–1.90 10/186,536 0.78 0.31–1.93 0.95 0.35–2.58
Total 22/186,329 1.00 Reference 1.00 Reference 12/186,257 1.00 Reference 1.00 Reference

Cancer Epidemiol Biomarkers Prev; 20(2) February 2011

isoflavonoidse
24/187,064 1.02 0.57–1.82 1.03 0.57–1.87 14/186,902 1.17 0.54–2.54 1.26 0.57–2.82
34/186,732 1.42 0.83–2.45 1.44 0.80–2.59 14/186,541 1.25 0.58–2.72 1.61 0.68–3.80
37/186,654 1.51 0.89–2.56 1.58 0.91–2.74 6/186,366 0.49 0.18–1.29 0.59 0.22–1.62
Coumestrol 66/469,675 1.00 Reference 1.00 Reference 33/469,317 1.00 Reference 1.00 Reference

Research.
22/139,351 1.11 0.68–1.84 1.16 0.67–1.85 10/139,229 1.50 0.71–3.16 1.56 0.74–3.33
29/137,752 1.41 0.91–2.18 1.47 0.94–2.29 3/137,521 0.29 0.09–0.94 0.32 0.10–1.07
Total fiber 27/185,299 1.00 Reference 1.00 Reference 16/185,167 1.00 Reference 1.00 Reference
28/187,158 0.95 0.55–1.61 0.94 0.55–1.62 12/187,027 0.72 0.34–1.53 0.75 0.35–1.62
24/187,476 0.76 0.44–1.32 0.72 0.40–1.30 10/187,315 0.55 0.25–1.22 0.59 0.25–1.37
38/186,845 1.13 0.68–1.85 1.06 0.59–1.89 8/186,558 0.37 0.16–0.87 0.41 0.16–1.09
Cereal fiber 23/186,018 1.00 Reference 1.00 Reference 12/185,904 1.00 Reference 1.00 Reference
35/186,701 1.49 0.88–2.53 1.49 0.87–2.53 14/186,488 1.27 0.59–2.76 1.32 0.61–2.89
22/187,233 0.90 0.50–1.62 0.88 0.49–1.61 11/187,136 0.98 0.43–2.23 1.05 0.45–2.43
37/186,827 1.44 0.85–2.42 1.37 0.79–2.39 9/186,539 0.70 0.30–1.67 0.74 0.30–1.85
Vegetable fiber 23/186,206 1.00 Reference 1.00 Reference 15/186,108 1.00 Reference 1.00 Reference
27/186,889 1.08 0.62–1.89 1.12 0.64–1.97 12/186,729 0.80 0.37–1.71 0.86 0.39–1.88
30/187,189 1.14 0.66–1.96 1.17 0.66–2.06 11/187,011 0.66 0.30–1.44 0.75 0.33–1.69
Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

37/186,495 1.31 0.77–2.22 1.35 0.76–2.39 8/186,219 0.40 0.17–0.94 0.50 0.20–1.27
a
Intake of Coumestrol is categorized into 3 groups, where the reference group contains those with zero intakes.
b
RR (95% CI) was obtained by Cox proportional hazards models.
c
Adjusted for age, oral contraceptives, age at menarche, parity, hormone replacement therapy, and intake of total energy intake, alcohol, saturated fat, meat, and fish.

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
d
Secoisolariciresinol, matairesinol, lariciresinol, isolariciresinol, pinoresinol, syringaresinol, and medioresinol.
e
Including genistein, daidzein, formononetin, biochanin A, and equol.

Cancer Epidemiology, Biomarkers & Prevention

 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Dietary Phytoestrogens and the Risk of Ovarian Cancer

Discussion study by McCann and ours are as follows: first, the

dietary sources of lignan differs—in the McCann study
In this large population-based prospective cohort the largest contributors were coffee, carrots, cucumber,
study, we found no evidence that specific phytoestrogens and strawberry, whereas in our study the lignan sources
or fiber intake are associated with overall ovarian cancer were rye bread, wheat bread, cereals, and berries. This
risk or ages less or more than 50 years. However, we difference is probably a result of a larger intake of bread
detected a difference in the effects of phytoestrogens or and cereals in our study population, but also due to that
fiber between invasive and borderline ovarian cancer risk. we included other lignan compounds than SECO and
We found a marginally inverse association between MAT in the total lignan intake. This is strengthening by
high intake of fiber or coumestrol (mainly reflecting bean the lack of association for grain products in the McCann
intake) and borderline ovarian cancer risk, but not for study. Second, the study by McCann was a case–control
invasive ovarian cancer risk. To our knowledge, such a study, a study design that is more vulnerable to certain
finding has not been reported before. However, the bias than cohort studies, for example, recall bias.
sample sizes were small, leading to unstable estimates. We found no association between intake of isoflavo-
Also, there are few studies on diet (non- on phytoestro- noids and ovarian cancer and this result is not in agree-
gens) and ovarian cancer, which has separated borderline ment with earlier findings of a protective effect of soy and
and invasive ovarian cancer, to compare our results with isoflavonoid intake on ovarian cancer risk (4). Most of the
and it is therefore difficult to rule out whether these are earlier studies were conducted in populations with a
chance findings. In a case–control study, an association higher intake of soy products, for example, in our cohort
was found between egg cholesterol and invasive ovarian the median value of the upper tertile of intake for bean,
cancer risk, but not for borderline ovarian cancer risk (35). soy, and lentil was 4.9 g/day, which is 10 times less than
Furthermore, in the same study, the observed inverse the lowest quartile of intake among women in a Chinese
relationship for vegetable fiber did not differ between case–control study (43). However, an inverse relationship
borderline or invasive cancer risks. In another case–con- between isoflavonoids and risk of ovarian cancer was
trol study, galactose consumption was associated with an found among Italian women (19), who have a lower
increased risk of borderline, but not invasive ovarian intake of isofalvonoids than Asian women. In the Italian
cancer (36). However, in case–control studies with retro- study, the inverse relationship was only seen for women
spective collected dietary data, it is difficult to rule out in the highest quintile of isofalvonoids intake, which was
whether dietary differences between invasive and bor- in fact less than the highest intake among women in our
derline cancer cases are due to etiologic differences or due cohort. The largest dietary contributors to the isoflavo-
to disease related changes as a result of a more sympto- noid intake in the Italian study was soy and soy milk (19),
matic stage of disease among invasive cancer cases. food items that is very uncommon in the Swedish diet,
However, results from the prospective US Nurses Health and the contributors to the isoflavonoid intake in our
study indicate that associations with several ovarian study was most likely other types of beans or vegetables.
cancer risk factors vary by histologic subtype, and these These differences in dietary habits could maybe be one of
differences were consistent with known similarities the reasons for lack of an association between isoflavo-
between each major histologic subtype and its normal noids and risk of ovarian cancer in our study, or again,
tissue counterpart (37). If tumors with low malignant differences in study designs.
potential (borderline ovarian tumors) are not considered In this study, we found no association between ovarian
as an intermediate precursor in the progression to high- cancer risk and phytoestrogen-rich food items such as
grade ovarian tumors, it is conceivable that invasive and berries, nuts, beans, crisp bread, or whole-grain bread.
borderline tumors may have different risk factors as well. For comparison, some other studies, investigating dietary
There is growing evidence that low- and high-grade pattern, have reported an increased risk for a starch-rich
serous carcinomas are distinctly different neoplasms with (44), "bread and pasta" (14), or high carbohydrate dietary
different pathogenesis, behavior, and response to treat- pattern (15), whereas others found no relationship with
ment (38, 39). Also, it is possible that the progression to ovarian cancer risk and intake of grain products (3).
high-grade tumors differs between the histologic sub- Furthermore, several studies have examined the role of
types of ovarian cancer (40–42). However, prospective vegetables and fruit for the risk of ovarian cancer, and
studies with sufficient power are needed to rule out if their results are inconsistent with some of them showing
exogenous factors influence low-grade tumors differently no association (10–13), an inverse association (3, 13, 14) or
than high-grade tumors. a positive association (15, 45). In addition, we found no
The lack of association between total lignans and ovar- association between risk of ovarian cancer and intake of
ian cancer risk in this study does not confirm earlier total carbohydrates, total cereals, or total vegetables.
findings from the only study that has investigated lignan Thus, the results in this study and current literature do
intake in relation to ovarian cancer risk (3). In this study, not provide any strong or consistent evidence for that
McCann and colleagues found a reduced risk of ovarian phytoestrogen-rich foods play a major role in ovarian
cancer among women with high intake of the lignan cancer development, at least among women with a low
compounds SECO and MAT. Discrepancy between the bean/soy intake.

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 20(2) February 2011 315

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Hedelin et al.

Comparison between studies of dietary phytoestrogen power to conclude, our results for crisp and whole-grain
intakes and disease is limited by differences in phytoestro- bread indicates that lignan containing foods are protec-
gen databases used in various studies. Databases often tive only among low consumers of antibiotics.
differ in calculation methods, chosen references, analytical We have only baseline dietary information, and mis-
methods, and origin of analysis of phytoestrogens in food classification of dietary intake maybe greater than if we
items. A strength of this study is that we used a phytoestro- had measured the diet at several points through the
gen database designed for a Swedish population (28) and follow-up time. However, the latency time for ovarian
that most of the phytoestrogen content in the foods were cancer is probably decades and our baseline measures
analyzed by 2 gas chromatography–mass spectrometric may reflects diet in the past, which may be the exposure
methods and carried out in the same laboratory (20). of interest.
Strengths of our study include its prospective design, In conclusion, we found no association between overall
large size, and complete follow-up. The ethnic homoge- ovarian cancer risk and intake of isoflavonoids, lignans,
neity of our study population (only 3% of the women and coumestrol, specific food items rich in phytoestro-
were born outside the northern countries) reduces the gens or fiber. Our overall results suggest that phytoestro-
risk of confounding by unmeasured factors. Cancer regis- gens does not play a major etiologic role in ovarian
tration in Sweden is obligatory, making the assessment of cancer, at least in this predominantly Caucasian women
cases almost complete. We adjusted for several known cohort with low bean/soy intake. However, the differ-
nutritional and nonnutritional risk factors for ovarian ences in the effect of diet between invasive and borderline
cancer. Also, we were able to take use of antibiotics into ovarian cancer need to be evaluated in other studies with
count, because antibiotics affect the bacterial microflora higher power.
in the gut (22) that transforms plant lignans into mam-
malian lignans (46); this could have an effect of phytoes- Disclosure of Potential Conflicts of Interest
trogens metabolism. However, the information on
No potential conflicts of interest were disclosed.
antibiotics was not available for the whole cohort, leading
to small sample size and unstable estimates. Further-
Acknowledgments
more, information about antibiotic use was based on
life-time consumption and we do not have any informa- The authors thank all study participants in the WLH study.
tion on the specific time of consumption. If antibiotic use
is distant, in time, from ovarian cancer development, Grant Support
antibiotic treatment may not have an impact on the
possible effect of phytoestrogens on ovarian cancer. How- This study was supported by Swedish Cancer Society Grant (06 0046),
ever, a large proportion (19%) of the women had used Swedish Research Council Grant (60433301), STINT (The Swedish Foundation
for International Cooperation in Research and Higher Education) Organon,
antibiotics more than 10 times in life and lignan metabo- Pharmacia, Medical Products Agency, and Schering-Plough. M. Hedelin was
lism could be affected by antibiotic use up to more than a partially supported by a postdoctoral fellowship from Folkhalsan Research
Center, Samfundet Folkhalsan, Helsinki, Finland.
year after consumption (22). Also, equol production is The costs of publication of this article were defrayed in part by the
reduced by antibiotics (23). Thus, at least for those payment of page charges. This article must therefore be hereby marked
women with frequent antibiotic use, the antibiotic treat- advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate
this fact.
ment could have influenced the production of bioactive
phytoestrogens in the gut and also influenced the meta- Received July 13, 2010; revised November 2, 2010; accepted November
bolism of the phytoestrogens. Even if we do not have the 17, 2010; published online February 7, 2011.

References
1. Adlercreutz H. Phyto-oestrogens and cancer. Lancet Oncol 2002;3: 7. Gossner G, Choi M, Tan L, Fogoros S, Griffith KA, Kuenker M, et al.
364–73. Genistein-induced apoptosis and autophagocytosis in ovarian cancer
2. Yan L, Spitznagel EL. Meta-analysis of soy food and risk of prostate cells. Gynecol Oncol 2007;105:23–30.
cancer in men. Int J Cancer 2005;117:667–9. 8. Makela SI, Pylkkanen LH, Santti RS, Adlercreutz H. Dietary soybean
3. McCann SE, Freudenheim JL, Marshall JR, Graham S. Risk of human may be antiestrogenic in male mice. J Nutr 1995;125:437–45.
ovarian cancer is related to dietary intake of selected nutrients, 9. Research WCRFAICR. Food, nutrition, physical activity, and the pre-
phytochemicals and food groups. J Nutr 2003;133:1937–42. vention of cancer: a global perspective. Washington, DC: AICR2007.
4. Myung SK, Ju W, Choi HJ, Kim SC. Soy intake and risk of endocrine- 10. Kolahdooz F, Ibiebele TI, Van Der Pols JC, Webb PM. Dietary patterns
related gynaecological cancer: a meta-analysis. BJOG 2009;116: and ovarian cancer risk. Am J Clin Nutr 2009;89:297–304.
1697–705. 11. Chang ET, Lee VS, Canchola AJ, Clarke CA, Purdie DM, Reynolds P,
5. Chen X, Anderson JJ. Isoflavones inhibit proliferation of ovarian et al. Diet and risk of ovarian cancer in the California Teachers Study
cancer cells in vitro via an estrogen receptor-dependent pathway. cohort. Am J Epidemiol 2007;165:802–13.
Nutr Cancer 2001;41:165–71. 12. Koushik A, Hunter DJ, Spiegelman D, Anderson KE, Arslan AA,
6. Gercel-Taylor C, Feitelson AK, Taylor DD. Inhibitory effect of genistein Beeson WL, et al. Fruits and vegetables and ovarian cancer risk in
and daidzein on ovarian cancer cell growth. Anticancer Res 2004;24: a pooled analysis of 12 cohort studies. Cancer Epidemiol Biomarkers
795–800. Prev 2005;14:2160–7.

316 Cancer Epidemiol Biomarkers Prev; 20(2) February 2011 Cancer Epidemiology, Biomarkers & Prevention

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Dietary Phytoestrogens and the Risk of Ovarian Cancer

13. Larsson SC, Holmberg L, Wolk A. Fruit and vegetable consumption in 31. Penalvo JL, Haajanen KM, Botting N, Adlercreutz H. Quantification of
relation to ovarian cancer incidence: the Swedish Mammography lignans in food using isotope dilution gas chromatography/mass
Cohort. Br J Cancer 2004;90:2167–70. spectrometry. J Agric Food Chem 2005;53:9342–7.
14. Edefonti V, Randi G, Decarli A, La Vecchia C, Bosetti C, Franceschi 32. Korn EL, Graubard BI, Midthune D. Time-to-event analysis of long-
S, et al. Clustering dietary habits and the risk of breast and ovarian itudinal follow-up of a survey: choice of the time-scale. Am J Epide-
cancers. Ann Oncol 2009;20:581–90. miol 1997;145:72–80.
15. Chang ET, Lee VS, Canchola AJ, Dalvi TB, Clarke CA, Reynolds P, 33. Willet W. Nutritional epidemiology. 2nd ed. New York: Oxford Uni-
et al. Dietary patterns and risk of ovarian cancer in the California versity Press; 1998.
Teachers Study cohort. Nutr Cancer 2008;60:285–91. 34. Weiderpass E, Baron JA, Adami HO, Magnusson C, Lindgren A,
16. Mueller SO, Simon S, Chae K, Metzler M, Korach KS. Phytoestrogens Bergstro€ m R, et al. Low-potency oestrogen and risk of endometrial
and their human metabolites show distinct agonistic and antagonistic cancer: a case-control study. Lancet 1999;353:1824–8.
properties on estrogen receptor alpha (ERalpha) and ERbeta in human 35. Risch HA, Jain M, Marrett LD, Howe GR. Dietary fat intake and risk of
cells. Toxicol Sci 2004;80:14–25. epithelial ovarian cancer. J Natl Cancer Inst 1994;86:1409–15.
17. Adlercreutz H. Lignans and human health. Crit Rev Clin Lab Sci 36. Cozen W, Peters R, Reichardt JK, Ng W, Felix JC, Wan P, et al.
2007;44:483–525. Galactose-1-phosphate uridyl transferase (GALT) genotype and phe-
18. Mazur W. Phytoestrogen content in foods. Baillieres Clin Endocrinol notype, galactose consumption, and the risk of borderline and inva-
Metab 1998;12:729–42. sive ovarian cancer (United States). Cancer Causes Control
19. Rossi M, Negri E, Lagiou P, Talamini R, Dal Maso L, Montella M, et al. 2002;13:113–20.
Flavonoids and ovarian cancer risk: a case-control study in Italy. Int J 37. Gates MA, Rosner BA, Hecht JL, Tworoger SS. Risk factors for
Cancer 2008;123:895–8. epithelial ovarian cancer by histologic subtype. Am J Epidemiol
20. Mazur W. Phytoestrogens: occurrance in foods, and metabolism of 2010;171:45–53.
lignans in man and pigs [dissertation]. University of Helsinki. 2000. 38. Vang R, Shih Ie M, Kurman RJ. Ovarian low-grade and high-grade
Available from:http://ethesis.helsinki.fi/julkaisut/laa/kliin/vk/mazur/. serous carcinoma: pathogenesis, clinicopathologic and molecular
21. Rose DP, Goldman M, Connolly JM, Strong LE. High-fiber diet biologic features, and diagnostic problems. Adv Anat Pathol 2009;16:
reduces serum estrogen concentrations in premenopausal women. 267–82.
Am J Clin Nutr 1991;54:520–5. 39. Vergara D, Tinelli A, Martignago R, Malvasi A, Chiuri VE, Leo G.
22. Kilkkinen A, Pietinen P, Klaukka T, Virtamo J, Korhonen P, Adlercreutz Biomolecular pathogenesis of borderline ovarian tumors: focusing
H. Use of oral antimicrobials decreases serum enterolactone con- target discovery through proteogenomics. Curr Cancer Drug Targets
centration. Am J Epidemiol 2002;155:472–7. 2010;10:107–16.
23. Setchell KD, Clerici C. Equol: history, chemistry, and formation. J Nutr 40. Huusom LD, Frederiksen K, Hogdall EV, Glud E, Christensen L,
2010;140:1355S-62S. Hogdal CK, et al. Association of reproductive factors, oral contra-
24. Lagiou P, Trichopoulos D, Sandin S, Lagiou A, Mucci L, Wolk A, et al. ceptive use and selected lifestyle factors with the risk of ovarian
Mediterranean dietary pattern and mortality among young women: a borderline tumors: a Danish case-control study. Cancer Causes
cohort study in Sweden. Br J Nutr 2006;96:384–92. Control 2006;17:821–9.
25. Hedelin M, Lof M, Olsson M, Adlercreutz H, Sandin S, Weiderpass E. 41. Jordan SJ, Green AC, Whiteman DC, Webb PM. Risk factors for
Dietary phytoestrogens are not associated with risk of overall breast benign, borderline and invasive mucinous ovarian tumors: epidemio-
cancer but diets rich in coumestrol are inversely associated with risk of logical evidence of a neoplastic continuum? Gynecol Oncol 2007;107:
estrogen receptor and progesterone receptor negative breast tumors 223–30.
in Swedish women. J Nutr 2008;138:938–45. 42. Jordan SJ, Green AC, Whiteman DC, Webb PM. Risk factors for
26. Wolk A, Bergstrom R, Hunter D, Willet W, Ljung H, Holmberg L, et al. A benign serous and mucinous epithelial ovarian tumors. Obstet Gyne-
prospective study of association of monounsaturated fat and other col 2007;109:647–54.
types of fat with risk of breast cancer. Arch Intern Med 1998;158:41–5. 43. Zhang M, Xie X, Lee AH, Binns CW. Soy and isoflavone intake are
27. National Food Administration, Uppsla, Sweden.[updated 2002; associated with reduced risk of ovarian cancer in southeast china.
cited 2005] Available from:http://www.slv.se/default.aspx?id = 231 Nutr Cancer 2004;49:125–30.
&epslanguage = EN-GB. 44. Edefonti V, Decarli A, La Vecchia C, Bosetti C, Randi G, Franceschi S,
28. Hedelin M, Klint A, Chang ET, Bellocco R, Johansson JE, Andersson et al. Nutrient dietary patterns and the risk of breast and ovarian
SO, et al. Dietary phytoestrogen, serum enterolactone and risk of cancers. Int J Cancer 2008;122:609–13.
prostate cancer: the cancer prostate Sweden study (Sweden). Cancer 45. Salazar-Martinez E, Lazcano-Ponce EC, Gonzalez Lira-Lira G, Escu-
Causes Control 2006;17:169–80. dero-De los Rios P, Hernandez-Avila M. Nutritional determinants of
29. Milder IE, Arts IC, van de Putte B, Venema DP, Hollman PC. Lignan epithelial ovarian cancer risk: a case-control study in Mexico. Oncol-
contents of Dutch plant foods: a database including lariciresinol, ogy 2002;63:151–7.
pinoresinol, secoisolariciresinol and matairesinol. Br J Nutr 2005; 46. Rowland IR, Wiseman H, Sanders TA, Adlercreutz H, Bowey EA.
93:393–402. Interindividual variation in metabolism of soy isoflavones and lignans:
30. Thompson LU, Boucher BA, Liu Z, Cotterchio M, Kreiger N. Phytoes- influence of habitual diet on equol production by the gut microflora.
trogen content of foods consumed in Canada, including isoflavones, Nutr Cancer 2000;36:27–32.
lignans, and coumestan. Nutr Cancer 2006;54:184–201.

www.aacrjournals.org Cancer Epidemiol Biomarkers Prev; 20(2) February 2011 317

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.
 Published OnlineFirst November 23, 2010; DOI: 10.1158/1055-9965.EPI-10-0752

Dietary Phytoestrogens and the Risk of Ovarian Cancer in

the Women's Lifestyle and Health Cohort Study
Maria Hedelin, Marie Löf, Therese M.-L. Andersson, et al.

Cancer Epidemiol Biomarkers Prev 2011;20:308-317. Published OnlineFirst November 23, 2010.

Updated version Access the most recent version of this article at:
doi:10.1158/1055-9965.EPI-10-0752

Cited articles This article cites 42 articles, 7 of which you can access for free at:
http://cebp.aacrjournals.org/content/20/2/308.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at:
http://cebp.aacrjournals.org/content/20/2/308.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Subscriptions Department at pubs@aacr.org.

Permissions To request permission to re-use all or part of this article, use this link
http://cebp.aacrjournals.org/content/20/2/308.
Click on "Request Permissions" which will take you to the Copyright Clearance Center's
(CCC)
Rightslink site.

Downloaded from cebp.aacrjournals.org on November 15, 2021. © 2011 American Association for Cancer
Research.