Content uploaded by Padmanabhan P Nair
Author content
All content in this area was uploaded by Padmanabhan P Nair on Feb 03, 2014
Content may be subject to copyright.
850
Am J Clin Nutr 1996;64:850-5. Printed in USA. © 1996 American Society for Clinical Nutrition
Effects of dietary fat and fiber on plasma and urine
androgens and estrogens in men: a controlled feeding
Joanne F Dorgan, Joseph T Judd, Christopher Longcope, Charles Brown, Arthur Schatzkin,
Beverly A Clevidence, William S Campbell, Padmanabhan P Nair, Charlene Franz, Lisa Kahie, and
Philip R Taylor
ABSTRACT We conducted a controlled feeding study to
evaluate the effects of fat and fiber consumption on plasma and
urine sex hormones in men. The study had a crossover design and
included 43 healthy men aged 19-56 y. Men were initially ran-
domly assigned to either a low-fat, high-fiber or high-fat, low-fiber
diet for 10 wk and after a 2-wk washout period crossed over to the
other diet. The energy content of diets was varied to maintain
constant body weight but averaged 13.3 MJ (3170 kcal)/d on
both diets. The low-fat diet provided 18.8% of energy from fat
with a ratio of polyunsaturated to saturated fat (P:S) of 1.3,
whereas the high-fat diet provided 41.0% of energy from fat with
a P:S of 0.6. Total dietary fiber consumption from the low- and
high-fat diets averaged 4.6 and 2.0 g - MJ ‘ - d ‘, respectively.
Mean plasma concentrations of total and sex-hormone-binding-
globulin (SHBG)-bound testosterone were 13% and 15% higher,
respectively, on the high-fat, low-fiber diet and the difference from
the low-fat, high-fiber diet was significant for the SHBG-bound
fraction (P = 0.04). Men’s daily urinary excretion of testosterone
also was 13% higher with the high-fat, low-fiber diet than with the
low-fat, high-fiber diet (P = 0.01). Conversely, their urinary
excretion of estradiol and estrone and their 2-hydroxy metabolites
were 12-28% lower with the high-fat, low-fiber diet (P 0.01).
Results of this study suggest that diet may alter endogenous sex
hormone metabolism in men. Am J Cliii Nutr l996;64:
850-5.
KEY WORDS Diet, dietary fats, dietary fiber, estrogen,
androgens
INTRODUCTION
Diet, particularly fat consumption, has been implicated as a
risk factor for prostate cancer. Prostate cancer mortality is
higher in countries where per capita consumption of animal fat
is elevated (1), and high-fat diets have been associated with an
increased risk of prostate cancer in several case-control and
cohort studies (2-9). In the Health Professionals Follow-up
Study, a prospective study of 51 529 men, the relative risk for
advanced prostate cancer was 1.79 (95% CI = 1.04, 3.07) for
those in the highest compared with the lowest quintile of fat
intake, and the association was primarily due to animal fat
consumption (8).
Prostate cancer is a hormone-dependent cancer and a
current hypothesis is that diet modifies prostate cancer risk
through an effect on the sex hormones (10, 1 1). Male
vegetarians have been reported to have lower plasma testos-
terone and estradiol concentrations than omnivores (1 2), and
dietary fat and fiber have been correlated with sex hormone
concentrations in several studies (12-14). Specific diet-
hormone relations reported in men include positive correla-
tions of testosterone with polyunsaturated fat (13) and di-
hydrotestosterone with vegetable fat consumption (14) and
inverse correlations of testosterone and estradiol with fiber
intake (1 2). However, comparisons between vegetarians and
omnivores and correlations of specific dietary components
with hormones from different studies have been inconsistent
(12-17). Moreover, in three diet intervention studies (18-
20), serum or urine testosterone levels were depressed with
the low-fat, high-fiber or vegetarian diet, but findings for
other hormones were inconsistent.
In 1986 the National Cancer Institute and Beltsville Human
Nutrition Research Center, Agriculture Research Service, con-
ducted a controlled feeding study in men to evaluate the effect
of modifying dietary fat and fiber intakes on several indexes
potentially related to cancer or atherosclerosis, including
plasma lipoproteins (21), prostaglandins (22), fecal mutagens,
and hormones. As part of this study, we evaluated the effect of
these dietary components on plasma and urine androgens and
estrogens.
1 From the Division of Cancer Prevention and Control, National
Cancer Institute, Bethesda, MD; the Diet and Human Performance
Laboratory, Beltsville Human Nutrition Research Center, ARS, US
Department of Agriculture, Beltsville, MD; the Departments of Obstet-
rics and Gynecology and Medicine, University of Massachusetts Mcd-
ical School, Worcester, MA; and Information Management Services,
Inc, Silver Spring, MD.
2 Address reprint requests to IF Dorgan, CPSA, DCPC, National Cancer
Institute, Executive Plaza North, Room 211, 6130 Executive Boulevard,
Bethesda, MD 20892-7326. E-mail: dorganj@dcpeepn.nci.nih.gov.
Received February 15, 1996.
Accepted for publication July 25, 1996.
by guest on July 15, 2011www.ajcn.orgDownloaded from
DIET AND ENDOGENOUS SEX HORMONES
851
SUBJECTS AND METHODS
Healthy male volunteers aged 19-56 y from the Beltsville,
MD, area who met the following criteria were recruited for the
controlled feeding study in 1986: 1) no history of diabetes,
cancer, or cardiovascular, kidney, or chronic gastrointestinal
disease; 2) no medication use other than an occasional analge-
sic; 3) a weight-for-height 80-130% of the desirable value
based on 1983 Metropolitan Life Insurance tables (23); 4)
normal results from a physical examination, including a com-
plete blood count and biochemical profile; and 5) no adherence
to a vegetarian diet in the past year. The research with human
volunteers was approved by the Institutional Review Boards of
Georgetown University and the National Cancer Institute. In-
formed consent was obtained from all men before enrollment.
The high-fat, low-fiber diet was designed to provide “‘40%
of energy from fat, with a ratio of polyunsaturated to saturated
fat (P:S) of 0.5, and 2.0 g total dietary fiber/Mi. The low-fat,
high-fiber diet was designed to provide < 20% of energy from
fat, a P:S of 1.2, and 4.6 g total dietary fiberfMJ. For both diets,
approximately one-third of the fiber was to come from each of
the fruit and vegetable, cereal, and legume groups. The study
used a crossover design. Participants were paired by age,
smoking status, and body mass index, and one man from each
pair was randomly assigned to each diet. After 10 wk on the
diet and a 2-wk washout period, participants were crossed over
to the other diet for 10 wk.
All meals were prepared in the Human Study Facility at the
Beltsville Human Nutrition Research Center. On weekdays
subjects ate breakfast and dinner in the dining room at this
facility and lunches were provided at breakfast for consump-
tion later in the day at work or at home. On weekends subjects
ate prepackaged meals at home. No foods other than those
provided by the study were permitted except tea and coffee.
Sweeteners and other additives to these beverages were limited
to those provided by the study, and quantities used were
recorded by participants. Salt was allowed ad libitum and
consumption was estimated by using tared salt shakers. Water
of known mineral content was provided to participants. Alco-
hol and vitamin and mineral supplements were prohibited.
Men were weighed on each weekday and the energy content
of the diets was varied in l.7-MJ (400-kcal) increments to
maintain constant body weight. Diets provided all known nu-
trients in amounts to meet recommended dietary allowances
(24). Nutrient composition was calculated by using US Depart-
ment of Agriculture (USDA) food-composition data together
with data from the food industry, the Nutrition Coordinating
Center at the University of Minnesota, and analyzed values. A
7-d menu cycle was used and composites of each menu were
analyzed to confirm the nutrient composition.
Blood was collected between 0600 and 0900 after a 12-h fast
on 1 d during the last week of each diet phase of the controlled
diet study. A 24-h urine sample was also collected on every day
of the final week and aliquots proportional to the total daily
volume from each day were pooled to create a single urine
specimen for each dietary period. Urine was collected on ice
and kept cold until measured and apportioned for storage. All
plasma and urine specimens were stored at - 80 #{176}Cor lower
until shipped to the laboratory on dry ice for hormone analyses,
which were performed between 1988 and 1989.
Testosterone, dihydrotestosterone, estrone, and estradiob
in plasma were measured by radioimmunoassay after sob-
vent extraction and celite chromatography (25). Androstan-
ediol glucuronide was measured by radioimmunoassay as
described by Horton et al (26-28). Dehydroepiandrosterone
sulfate (DHEAS) was measured with a radioimmunoassay
kit (ICN-Biomedical, Costa Mesa, CA) and SHBG was
measured with an immunoradiometric assay kit (Farmos
Group Ltd, Oulunsalo, Finland). The percentages of un-
bound and albumin-bound testosterone and estradiol were
measured by using centrifugal ultrafiltration (29, 30), and
percentages of SHBG-bound hormones were calculated.
Urinary testosterone glucuronide was measured with a ra-
dioimmunoassay kit (ICN-Biomedical) and our results were
similar to those reported for normal males by Doberne and
New (3 1 ) and Tresguerres et al (32) after hydrolysis with
-glucuronidase. The urinary estrogens were measured by
radioimmunoassay after 3-glucuronidase hydrolysis (33)
and LH-20 sephadex chromatography (Pharmacia LKB Bio-
technology, Piscataway, NJ) as described (34). Urinary
2-hydroxyestrone and 2-hydroxyestradiol were measured by
radioimmunoassay according to the methods of Chattoraj et
al (35).
For each analyte, plasma and urine samples from the same
man were analyzed in the same batch to remove the effects of
between-batch variability in hormone assays. Within-batch
CVs of plasma hormone measurements in replicate quality-
control samples averaged 8.2% for estrone, 5.7% for estradiol,
11.8% for testosterone, 13.7% for dihydrotestosterone, and
8.9% for DHEAS. The within-batch CV for the percentage free
and albumin-bound estradiol and testosterone were all < 10%
except for free estradiol, which was 10.8%. In urine the within-
batch CV for glucuronides averaged 16.0% for estrone, 6.8%
for estradiob, 20.8% for estriol, and 9.2% for testosterone. For
the catechols, the within-batch CV averaged 9.4% for 2-hy-
droxyestrone and 12.0% for 2-hydroxyestradiol.
Because distributions of plasma and urine hormones were
not normal, geometric means were used to describe the data.
Methods proposed by Fleiss (36) were used to evaluate
carryover and diet effects. To determine whether carryover
from period 1 to period 2 was the same regardless of initial
diet, for each hormone the sum of levels from both periods
for men initially randomly assigned to the high-fat, low-
fiber diet were compared with the sum for men initially
randomly assigned to the low-fat, high-fiber diet by using a
Wilcoxon rank-sum test. Because there was no evidence of
differential carryover for any hormone at a significance
level of 0.10, carryover was ignored when the effect of
diet was evaluated. The diet effect was estimated by the
difference between hormone levels after the high-fat, low-
fiber and low-fat, high-fiber diets regardless of diet order.
The significance of the diet effect was tested by comparing
the difference in hormone levels between periods 1 and 2 for
men initially randomly assigned to the high-fat, low-fiber
diet with the same difference for men initially randomly
assigned to the low-fat, high-fiber diet using a t test if the
distribution of differences was normally distributed and a
Wilcoxon rank-sum test otherwise. All analyses were per-
formed by using SAS statistical software (37).
by guest on July 15, 2011www.ajcn.orgDownloaded from
852
DORGAN ET AL
I Ratio of polyunsaturated to saturated fat.
RESULTS
DISCUSSION
Of the 45 men randomly assigned to the controlled diet
study, 43 completed both phases. Their mean (± SD) age was
33.8 ± 9.2 y and their median body weights at baseline and at
the end of the controlled feeding study, respectively, were 79.6
kg (range: 58.2-122.2 kg) and 79.0 kg (range: 60.2-1 15.8 kg),
which were not significantly different. Seven (16%) of the men
were black and 10 (23%) smoked cigarettes.
The nutrient composition of the actual diets consumed dur-
ing each study period is shown in Table 1. Energy intake was
= 13.3 MJ (3 170 kcal)/d on both the high-fat, low-fiber and
low-fat, high-fiber diets. The median daily percentage of en-
ergy from fat was 41.0% (range: 38.6-42.0%) with the high-fat
diet and 18.8% (range: 17.4-20.5%) with the low-fat diet. Type
of fat also differed between diets; the median P:S was 0.6
(range: 0.5-0.7) for the high-fat diet compared with 1 .3 (range:
1.0-1.6) for the low-fat diet. Total dietary fiber intake from the
low-fat diet was more than twice that from the high-fat diet.
Hormones in plasma and urine specimens collected at the
end of each period of the controlled diet study and differences
between the high-fat, low-fiber and low-fat, high-fiber diets are
shown in Tables 2 and 3. Plasma concentrations of androgens
tended to be elevated with the high-fat, low-fiber diet; geomet-
ric mean concentrations of total and SHBG-bound testosterone
were 13% and 15% higher, respectively, compared with the
bow-fat, high-fiber diet and for SHBG-bound testosterone the
mean difference was significant (P = 0.04). Men’s daily cx-
cretion of testosterone glucuronide was also 13% higher when
the high-fat, low-fiber diet was consumed than when the bow-
fat, high-fiber diet was consumed, and the mean difference was
significant (P = 0.01). We also evaluated the effect of diet on
the ratio of dihydrotestosterone to testosterone in plasma and
found that the ratio was essentially unchanged after the low-fat,
high-fiber diet compared with that after the high-fat, low-fiber
diet; the mean difference was -0.01 (95% CI = -0.02, 0.01).
Trends were inconsistent for the plasma estrogens, and none
of the differences examined were significant. However, men’s
excretion of glucuronides of estradiol and estrone and their
2-hydroxy metabolites were all 12-28% bower after the high-
fat, low-fiber diet (P 0.01).
In this controlled feeding study plasma androgens in a single
blood draw tended to be elevated when men ate a high-fat,
low-fiber compared with a low-fat, high-fiber diet for 10 wk.
With the high-fat, low-fiber diet, amounts of testosterone glu-
curomde were also greater and estradiob and estrone glucu-
ronide amounts were bower in 24-h urine samples collected
over 1 wk and pooled.
Howie and Shultz (12) reported lower plasma testosterone
and estradiol concentrations in male vegetarians than in meat
eaters, but the majority of studies that compared plasma an-
drogens and estrogens in vegetarians and omnivores did not
detect differences (13, 15-17). Vegetarians were reported to
have elevated SHBG concentrations in two studies (13, 16),
and in one (16) the ratio of testosterone to SHBG was de-
pressed, suggesting that vegetarians may have less non-SHBG-
bound, or bioavailabbe, testosterone compared with omnivores.
In the study by Howie and Shultz (12), plasma testosterone and
estradiol concentrations were inversely correlated with dietary
fiber ingestion. Key et al (13) on the other hand found positive
correlations of testosterone with polyunsaturated fat intake and
SHBG with total, saturated, and polyunsaturated fat intakes.
Twenty-four-hour excretions of the androgens DHEAS, an-
drosterone, and etiocholanobone, and the estrogens estradiol,
estrone, and estriol were significantly lower in middle-aged
South African blacks, who customarily follow a vegetarian diet
compared with North American blacks eating meat (18). When
the North Americans were fed a diet without any meat or meat
products, their 24-h urine excretion of androgens and estrogens
decreased significantly, and when the South Africans were
switched to a Western diet including meat, their urine output of
these hormones increased. Diet-hormone relations were age-
dependent, however, and when South African blacks aged
60 y ate meat, their urine estrogens and androgens and serum
androgens were lower than when they consumed a vegetarian
diet (18, 38).
Because vegetarians could differ from omnivores on char-
acteristics other than diet that influence hormone concentra-
tions, Raben et al (20) investigated the effect of consuming a
vegetarian diet on serum hormones in a controlled study. Eight
men were fed a lactoovovegetarian diet and a mixed-meat diet
TABLE 1
Medians and ranges of daily nutrient consumption by diet
High-fat, low-fiber diet
Low-fat, high-fiber diet
Median Range Median Range
Energy (MJ) 13.2
10.2-18.5
13.3 10.0-18.5
Protein(%ofenergy)
14.8
13.0-18.1 17.1
16.0-17.9
Carbohydrate (% of energy)
45.3
43.6-48.8 67.5 66.0-68.7
Total fat (% of energy) 41.0 38.6-42.0
18.8 17.4-20.5
Saturated fat (% of energy) 14.7
13.9-15.4
4.4 3.7-5.2
Linoleic acid (% of energy) 8.1 6.6-8.8
5.3 4.1-6.3
Oleic acid (% of energy)
14.1
12.1-16.1
6.4 5.5-7.4
P:S’ 0.6
0.5-0.7 1.3 1.0-1.6
Cholesterol (mgIMJ)
45.1 44.1-46.3
18.2 17.5-18.6
Total dietary fiber (g/MJ)
2.0 2.0-2.0
4.6 4.6-4.6
by guest on July 15, 2011www.ajcn.orgDownloaded from
DIET AND ENDOGENOUS SEX HORMONES 853
TABLE 2
Geometric mean (95% CI) concentrations and mean (95% CI) differences in plasma sex-hormone-binding globulin (SHBG), androgens, and estrogens
by diet
High-fat , low-fiber diet Low-fat, high-fiber diet High-fat diet - low-fat diet
Mean 95% CI Mean 95% CI Difference’
95% CI P
SHBG (nmol/L) 19.2 16.9, 21.6 18.5 16.2, 21.2 1.2
-0.6, 2.9 0.33
Androgens
Testosterone (nmolIL)
Total 13.3
11.6, 15.3 11.8 10.1, 13.8 1.6 -0.2,3.5 0.10
Free 0.31 0.26, 0.36 0.33 0.23, 0.28
0.03 -0.02, 0.08 0.27
Albumin-bound 4.4 3.8, 5.2 4.1 3.4, 4.9 0.4 -0.4, 1.1 0.37
SHBG-bound 8.2 7.0, 9.7
7.1 6.0, 8.5 1.2 0.1, 2.4 0.04
Dihydrotestosterone (nmolIL) 1.2
1.1, 1.4 1.1 0.9, 1.3 0.09
-0.04, 0.22 0.20
DHEAS (mol/L) 9.3 8.2, 10.5 8.8 7.6, 10.1
0.4 -0.6, 1.4 0.10
Androstanediol glucuronide (nmoLfL) 1.7 1.5, 1.9 1.7 1.5, 1.9 0.1 -0.1, 0.2 0.99
Estrogens
Estrone (pmolIL)
154.1 140.3, 170.0 148.0 133.8, 164.7 4.8 -6.3, 16.0 0.28
Estradiol (pmol/L)
Total 99.4 87.9, 112.6 104.6 90.9, 120.3 -6.8 -18.7,5.0 0.31
Free 2.2 2.0, 2.5 2.4 2.1, 2.8 -0.3 -0.6, 0.1 0.15
Albumin-bound
32.9 28.8, 37.4 36.3 30.7, 42.7 - 1.7 -6.1, 2.3 0.12
SHBG-bound 63.4 55.7, 72.2 64.9 56.5, 74.1 -2.9 - 10.3, 4.4 0.46
‘ Mean differences of untransformed values after removal of outliers: total estradiol (n 1), free estradiol (n I), albumin-bound estradiol (n 3).
DHEAS, dehydroepiandrosterone sulfate.
2 p value for test of H0: difference = 0 when using all observations based on a t test when differences were normally distributed (total, free,
albumin-bound, and SHBG-bound testosterone; dihydrotestosterone; and SHBG-bound estradiol) and a Wilcoxon rank-sum test when differences were not
normally distributed (SHBG, DHEAS, androstanediol glucuronide, estrone, and total, free, and albumin-bound estradiol).
each for 6 wk. The diets were isoenergetic and “‘28% of
energy was derived from fat on both diets. The P:S, however,
was > 1 with the vegetarian diet compared with “0.5 with the
meat diet and the fiber content of the vegetarian diet was
approximately twice that of the meat diet. Predict serum con-
centrations of testosterone were comparable and decreased
significantly by 35% after the vegetarian diet but not after the
meat diet. Differences in other androgens and estrogens were
also reported, but these were attributed to dissimilarities in
baseline concentrations.
Hamabainen et al (19) studied the effect of modifying dietary
fat on serum sex hormones in 30 healthy, free-living men
40-49-y old. Serum sex hormones were measured after a 2-wk
period when men consumed their usual diets, which provided
40% of energy as fat with a P:S of 0.15 and again after a 6-wk
intervention period when men consumed isoenergetic diets that
provided 25% of energy as fat with a P:S of 1.22. After the
intervention, men had significantly lower serum concentrations
of androstenedione and total and free testosterone. However,
concentrations of dihydrotestosterone, DHEAS, estradiol, and
estrone did not differ between the two periods.
Reed (39) failed to detect a difference in total testosterone in
six men following isoenergetic diets providing 20 and 100 g
fat/cl. Because SHBG concentrations increased, however, free
testosterone decreased after the low-fat diet.
Our finding of greater daily urine testosterone excretion in
men consuming a high-fat, low-fiber diet relative to men con-
suming a low-fat, high-fiber diet is consistent with the findings
of most controlled feeding studies. We also observed elevated
plasma testosterone concentrations when men ate the high-fat,
low-fiber diet, but the difference from concentrations in men
who consumed the low-fat, high-fiber diet was significant only
TABLE 3
Geometric mean (95% CI) amounts and mean (95% CI) differences in daily urine hormones by diet
High-fat , low-fiber diet Low-fat,
high-fiber diet High- fat diet - low-fat diet
Mean 95% Cl
Difference’ 95% CI P
Mean 95% CI
Glucuronides
Testosterone (nmol/d) 161.3 124.6, 207.4 143.0
1 12.5, 180.8 26.1 9.6, 42.6 0.01
Estrone (nmol/d) 1 1.1 9.8, 12.6 13.8 12.3, 15.6 -2.4 -3.4, - 1.3 0.0002
Estradiol (nmolld) 3.8 3.4, 4.3 4.3 3.7, 4.9 -0.6 - 1.0, -0.2
0.008
Estriol (nmol/d) 6.8 5.5, 8.3 6.8 5.5, 8.4 0.3 -0.7, 1.4 0.57
Catechols
2-hydroxyestrone (nmol/d) 71.8 65.1, 78.6
85.2 76.4, 94.1 - 14.6 -21.6, -7.5 0.0003
2-hydroxyestradiol (nmolld)
9.3 8.1, 10.7 12.9 11.5, 14.4
-3.4 -4.8, -2.0 <0.0001
I Mean differences of untransformed values after removal of outliers: testosterone (n = 2), estrone (n 1), estriol (n 1).
2 p value for test of H0: difference = 0 when using all observations based on a I test when differences were normally distributed (estradiol,
2-hydroxyestrone, 2-hydroxyestradiol) and a Wilcoxon rank-sum test when differences were not normally distributed (testosterone, estrone, and estriol).
by guest on July 15, 2011www.ajcn.orgDownloaded from
854
DORGAN ET AL
for the fraction that was SHBG-bound. The lack of significance
for plasma total testosterone may have been due to within-
person variation in hormone concentrations. Plasma testoster-
one has been reported to exhibit considerable diurnal variation,
falling 42% from morning until night in one study (40). Fur-
thermore, because of short-term fluctuations in blood concen-
trations, Goldzieher et al (41) estimated that a testosterone
value from a single blood draw would be within 20% of the
underlying mean only 68% of the time. Testosterone amounts
in 24-h urine specimens collected over 1 wk and pooled,
therefore, are probably a more valid assessment than are con-
centrations in a single-morning blood draw collected after an
overnight fast.
We observed lower daily urine estradiol and estrone glu-
curonide excretion in men after consumption of a high-fat,
low-fiber diet. Hill et al ( 18) also reported lower urine
estrogens in South African black men aged 60 y who ate
a diet with meat rather than their usual vegetarian diet. In
younger men aged < 55 y, the opposite effect was observed,
indicating that age modified the relation of diet and urine
hormones. The men in our study were 19-56 y of age; when
we analyzed the data separately for those aged 30 y and
for those aged > 30 y, both groups had lower urine estro-
gens after consuming the high-fat, low-fiber diet.
Because the major source of estrogen in men is the pe-
ripheral aromatization of androgens, our finding of an in-
crease in urinary excretion of testosterone but a slight de-
crease in the excretion of estrone and estradiol after a
high-fat low-fiber diet was unexpected. However, it is pos-
sible that this was a result of a decrease in the peripheral
aromatization of androgens that offset the slight increase in
androgen amounts. Although peripheral aromatization of
androgens is influenced by body weight (42), the men’s
weight was stable throughout the study. No data are avail-
able on the effect of dietary fat and fiber on peripheral
aromatization.
The major urinary metabolites of estradiol are the
catechols (2-hydroxyestradiol and 2-hydroxyestrone) and
16a-hydroxyestrone. Although controversial (43), 16a-hy-
droxyestrone has been implicated in mammary carcinogen-
esis (44) and could potentially play a robe at different sites.
We observed decreased excretion of catechol estrogens after
a high-fat, low-fiber diet. Although we did not measure
urinary 16a-hydroxyestrone, we did measure estriol, which
was unaffected by diet in our study. Among women, Long-
cope et al (45) found that consumption of a high-fat diet
resulted in decreased excretion of catechol estrogens and
significantly increased excretion of l6a-hydroxyestrone
and its metabolite estriol. Adlercreutz et al (43), however,
reported no difference in urinary of 2-hydroxyestrone be-
tween omnivorous and vegetarian women, although their
dietary fat and fiber intakes differed significantly. Addi-
tional studies are needed to clarify whether discrepancies in
the effects of dietary fat and fiber on the pathways of
estrogen metabolism reflect a sex difference and/or are due
to differences in study design.
In summary, results of this controlled feeding study suggest
that dietary fat and fiber may affect sex hormone metabolism in
men in a way that may influence prostate cancer risk. CI
REFERENCES
1. Rose DP, Boyar AP, Wynder EL. International comparisons of mor-
tality rates for cancer of the breast, ovary, prostate, and colon, and per
capita food consumption. Cancer 1986;58:2363-71.
2. Graham 5, Haughey B, Marshall J, et al. Diet in the epidemiology of
carcinoma of the prostate gland. J Nail Cancer Inst 1983;70:687-92.
3. Ross RK, Shimizu H, Paganini-Hill A, Honda G, Henderson BE.
Case-control studies of prostate cancer in blacks and whites in South-
em California. J Natl Cancer Inst 1987;78:869-74.
4. Kolonel L, Yoshizawa CN, Hankin JH. Diet and prostatic cancer: a
case-control study in Hawaii. Am J Epidemiol l988;l27:999-lOl2.
5. West DW, Slattery ML, Robison LM, French TK, Mahoney AW.
Adult dietary intake and prostate cancer risk in Utah: a case-control
study with special emphasis on aggressive tumors. Cancer Causes
Control 1991 ;2:85-94.
6. Bravo MP, Castellanos E, del Rey Calero J. Dietary factors and
prostatic cancer. Urol Int 1991 ;46: 163-6.
7. Whittemore AS, Kolonel LN, Wu AH, et al. Prostate cancer in relation
to diet, physical activity, and body size in blacks, whites, and Asians
in the United States and Canada. J Natl Cancer Inst 1995;87:652-61.
8. Giovannucci E, Rimm EB, Colditz GA, et al. A prospective study of
dietary fat and risk of prostate cancer. J Natl Cancer Inst
l993;85: 1571-9.
9. Le Marchand L, Kolonel LN, Wilkens LR, Myers BC, Hirohata T.
Animal fat consumption and prostate cancer: a prospective study in
Hawaii. Epidemiology 1994;5:276-82.
10. Pienta K, Esper PS. Is dietary fat a risk factor for prostate cancer?
J Nail Cancer Inst 1993;85: 1538-40.
1 1. Ross RK, Henderson BE. Do diet and androgens alter prostate cancer
risk via a common etiologic pathway? J Natl Cancer Inst
l994;86:252-4.
12. Howie BJ, Shultz TD. Dietary and hormonal interrelationships among
vegetarian Seventh-Day Adventists and nonvegetarian men. Am J Clin
Nutr l985;42: 127-34.
13. Key TJA, Roe L, Thorogood M, Moore JW, Clark GMG, Wang DY.
Testosterone, sex hormone-binding globulin, calculated free testoster-
one, and oestradiol in male vegans and omnivores. Br J Nutr
1990;64:l 11-9.
14. Field AE, Colditz GA, Willert WC, Longcope C, McKinlay JB. The
relation of smoking, age, relative weight, and dietary intake to serum
adrenal steroids, sex hormones, and sex hormone-binding globulin in
middle-aged men. J Clin Endocrinol Metab 1994;79: 1310-6.
15. Deslypere JP, Vermeulen A. Leydig cell function in normal men:
effect of age, life-style, residence, diet, and activity. J Clin Endocrinol
Metab 1984;59:955-62.
16. Belanger A, Locong A, Noel C, et al. Influence of diet on plasma
steroid and sex plasma binding globulin levels in adult men. J Steroid
Biochem l989;32:829-33.
17. Pusateri DJ, Roth WT, Ross JK, Shultz TD. Dietary and hormonal
evaluation of men at different risks for prostate cancer: plasma and
fecal hormone-nutrient interrelationships. Am J Clin Nutr
1990;51 :371-7.
18. Hill P, Wynder EL, Garbaczewski L, Games H, Walker ARP. Diet and
urinary steroids in black and white North American men and black
South African men. Cancer Res l979;39:SlOl-5.
19. Hamalainen EK, Adlercreutz H, Puska P. Pietinen P. Diet and serum
sex hormones in healthy men. J Steroid Biochem 1984;20:459-64.
20. Raben A, Kiens B, Richter EA, et al. Serum sex hormones and
endurance performance after a lactoovovegetarian and a mixed diet.
Med Sci Sports Exerc l992;24: 1290-7.
21. Clevidence BA, Judd JT, Schatzkin A, et al. Plasma lipid and lipopro-
tein concentrations of men consuming a low-fat, high-fiber diet. Am J
Clin Nutr 1992;55:689-94.
22. Ferretti A, Judd if, Taylor PR, Schatzkin A, Brown C. Modulating
influence of dietary lipid intake on the prostaglandin system in adult
men. Lipids 1989;24:4l9-22.
by guest on July 15, 2011www.ajcn.orgDownloaded from
DIET AND ENDOGENOUS SEX HORMONES 855
23. Metropolitan Life Insurance Company. 1983 Height and weight tables.
Stat Bull Metropol Insur Co 1983;64:2-9.
24. National Research Council. Recommended dietary allowances. 9th ed.
Washington, DC: National Academy of Sciences, 1980.
25. Longcope C, Franz C, Morello C, Baker R, Johnston CC. Steroid and
gonadotropin levels in women during the peri-menopausal years. Ma-
turitas 1986;8:l89-96.
26. Horton R, Hawks D, Lobo R. 3a,l73-androstanediol glucuronide in
plasma: a marker of androgen action in idiopathic hirsutism. J Clin
Invest l982;69: 1203-6.
27. Barberia J, Pages L, Horton R. Measurement of androstanediol in
plasma in a radioimmunoassay using celite column chromatography.
Fertil Steril l976;27: 1101-4.
28. Horton R, Endres D, Galmarini M. Ideal conditions for hydrolysis of
androstane-3a,l73-diol glucuronide in plasma. J Clin Endocrinol
Metab l986;62:22-7.
29. Longcope C, Femino A, Johnston JO. Androgen and estrogen dynam-
ics in the female baboon (Papio anubis). J Steroid Biochem
l988;3l :195-200.
30. Longcope C, Hui SL, Johnston CC. Free estradiol, free testosterone
and sex hormone binding globulin in pen-menopausal women. J Clin
Endocrinol Metab 1987;64:5l3-8.
31. Doberne Y, New MI. Urinary androstanediol and testosterone in
adults. J Clin Endocrinol Metab 1976;42: 152-4.
32. Tresguerres JA, Lisboa BP, Tamm J. A simple radioimmunoassay for
the measurement of testosterone glucosiduronate in unextracted urine.
Steroids 1976;28: 13-23.
33. Longcope C. Methods and results of aromatization studies in vivo.
Cancer Res 1982;42(suppl):3307s-l ls.
34. Rotti K, Stevens J, Watson D, Longcope C. Estriol concentrations in
plasma of normal, non-pregnant women. Steroids l975;25:807-l6.
35. Chattoraj SC, Fanous AS, Cecchini D, Lowe EW. A radioimmunoas-
say method for urinary catechol estrogens. Steroids 1978;31:375-9l.
36. Fleiss IL. The design and analysis of clinical experiments. New York:
John Wiley and Sons, 1986:263-90.
37. SAS Institute, Inc. SAS user’s guide, version 5. Cary, NC: SAS
Institute Inc. 1985.
38. Hill P, Wynder EL, Garbaczewski L, Walker ARP. Effect of diet on
plasma and urinary hormones in South African black men with pros-
tatic cancer. Cancer Res 1982;42:3864-9.
39. Reed MJ, Cheng RW, Simmonds M, Richmond W, James VHT.
Dietary lipids: an additional regulator of plasma levels of sex hormone
binding globulin. J Clin Endocrinol Metab 1987;64:l083-5.
40. Rose RM, Kreuz LE, Holaday JW, Sulak KJ, Johnson CE. Diurnal
variation of plasma testosterone and cortisol. J Clin Endocrinol Metab
1972;54: 177-8.
41. Goldzieher JW, Dozier TS, Smith KD, Steinberger E. Improving the
diagnostic reliability of rapidly fluctuating plasma hormone levels by
optimized multiple-sampling techniques. J Clin Endocrinol Metab
l976;43:824-30.
42. Simpson ER, Merrill JC, Alexander J, Hollub AJ, Graham-Lorence 5,
Mendelson CR. Regulation ofestrogen biosynthesis by human adipose
cells. Endocrine Rev l989;l0:136-48.
43. Adlercreutz H, Fotsis T, Hockerstedt K, et al. Diet and urinary estro-
gen profile in premenopausal omnivorous and vegetarian women and
in premenopausal women with breast cancer. J Steroid Biochem
l989;34:527-30.
44. Fishman J, Telang NT. The role of estrogen in mammary carcinogen-
esis. Ann NY Acad Sci 1995;768:91-lOO.
45. Longcope C, Gorbach S. Goldin B, et al. The effect of a low fat diet
on estrogen metabolism. J Clin Endocrinol Metab l987;64:1246-50.
by guest on July 15, 2011www.ajcn.orgDownloaded from
