NUTRITION AND CANCER, 58(1), 35–42
Copyright C ?2007, Lawrence Erlbaum Associates, Inc.
Relationship of Dietary Protein and Soy Isoflavones to Serum IGF-1
and IGF Binding Proteins in the Prostate Cancer Lifestyle Trial
Antonella Dewell, Gerdi Weidner, Michael D. Sumner, R. James Barnard, Ruth O. Marlin,
Jennifer J. Daubenmier, Christine Chi, Peter R. Carroll, and Dean Ornish
Abstract: High levels of insulin-like growth factor 1 (IGF-
1) are associated with increased risk of prostate cancer,
whereas increased levels of some of its binding proteins
(IGFBPs) seem to be protective. High intakes of dietary
protein, especially animal and soy protein, appear to in-
crease IGF-1. However, soy isoflavones have demonstrated
anti-proliferative and apoptotic effects both in vitro and
in vivo. We evaluated dietary intakes of total protein and
soy isoflavones in relation to the IGF axis in prostate can-
cer patients making comprehensive lifestyle changes in-
cluding a very low-fat vegan diet supplemented with soy
protein (58 g/day). After one year, intervention group pa-
tients reported significantly higher intakes of dietary protein
and soy isoflavones compared to usual-care controls (P <
IGFBP-1 rose in the experimental group only (P < 0.01). In-
creases in vegetable protein over one year were associated
with increases in IGFBP-1 among intervention group pa-
tients (P < 0.05). These results suggest that dietary protein
and soy isoflavones, in the context of comprehensive lifestyle
changes, may not significantly alter IGF-1. However, given
IGF-1, it may be prudent for men with early stage prostate
cancer not to exceed dietary protein recommendations.
Insulin-like growth factor 1 (IGF-1),a hormone that plays
been shown to promote tumor growth and inhibit apoptosis
(1,2). A growing body of epidemiological literature suggests
that individuals with high IGF-1 levels are at increased risk
of various types of cancer, including prostate cancer (3), and
especially advanced-stage prostate cancer (4). Conversely,
there is some evidence that two of the main IGF-1 binding
Antonella Dewell, Gerdi Weidner, Michael D. Sumner, Ruth O. Marlin, Christine Chi, and Dean Ornish are affiliated with Preventive Medicine Research
Institute, Sausalito, California. James Barnard is affiliated with University of California, Los Angeles, California. Jennifer J. Daubenmier and Peter R. Carroll
are affiliated with University of California, San Francisco, California.
proteins (IGFBPs), IGFBP-1 and IGFBP-3, may be protec-
tive (4–8). In fact, individuals with both the highest levels
of IGF-1 and the lowest levels of IGFBP-3 may be those at
highest risk (4).
The major determinants of the IGF-1 axis include both
genetic and lifestyle factors (e.g., diet and exercise), and age
(1,2). Several observational and human feeding studies have
identified protein as the most important dietary factor influ-
protein rich in essential amino acids (animal or soy protein)
are thought to be responsible for the observed increases in
IGF-1 (9,12,15). In addition, IGF-1 levels are known to de-
crease with interventions including a very low-fat diet and/or
exercise (5,6). IGF-1 levels are known to increase from birth
to puberty and to decline with age afterwards. These changes
are regulated by growth hormone. Another important deter-
minant of circulating IGF-1 is insulin, which stimulates its
Recent research investigating the dietary determinants of
IGF-1 has focused on soy protein. The use of soy in Asian
diets is theorized to be one of the reasons for the low in-
cidence and mortality rates from prostate cancer in Asian
men, which are rapidly increasing with the Westernization
of their traditional diet (17). This hypothesis is supported by
findings from both laboratory and animal studies, indicat-
ing an anti-proliferative and apoptotic effect in cancer cells
of isoflavones, the phytoestrogens present in soy (18–21).
However, epidemiological and clinical investigations on the
role of soy protein in relation to IGF-I suggest that soy pro-
tein may alter the IGF axis toward an increase in IGF-1 and
a reduction in IGFBP-3 (22–26).
We present a post-hoc analysis of the relationship be-
tween dietary intakes of total protein and soy isoflavones to
IGF-1, its main binding proteins, IGFBP-1, IGFBP-2, and
IGFBP-3, and the IGF-1: IGFBP-3 molar ratio (an esti-
mate of free, i.e., available, IGF-1) in the Prostate Cancer
Lifestyle Trial (PCLT). In this randomized, controlled trial,
intervention group participants enrolled in a multicompo-
nent lifestyle intervention including a very low-fat vegan
diet supplemented with soy protein, exercise, stress manage-
ment, and group support, and were compared to a usual-care
control group (27–29). Specifically, our focus is to evaluate
whether high dietary intakes of protein and soy isoflavones
with increases in serum IGF-1. Additionally, we investigate
associations of dietary protein, soy isoflavones, and exercise,
with prostate cancer markers (i.e., prostate specific antigen
(PSA), serum-stimulated LNCaP cell growth and apoptosis,
and testosterone) and with fasting insulin.
Subjects and Methods
The subjects were 93 men participating in the PCLT, a
randomized controlled study investigating whether compre-
hensive diet and lifestyle changes may affect the progres-
sion of prostate cancer (27). Participants included men with
domized to a lifestyle intervention (n = 44) or usual-care
control group (n = 49). The University of California San
Francisco Committee on Human Research Institutional Re-
previously reported (27–29). Briefly, patients making inten-
sive lifestyle changes had significantly lower PSA levels and
in vitro serum-stimulated LNCaP cell growth compared to
patients in the control group (27).
procedure, intervention, and assessment protocol have been
tervention group were asked to follow an intensive lifestyle
program including a very low-fat vegan diet (approximately
10% energy from fat), moderate aerobic exercise (e.g., walk-
ing, meditation, imagery, and progressive relaxation for a to-
tal of 60 min daily), and social group support (1 h weekly).
of a soy product and a 58-g serving of a fortified soy protein
powdered beverage (SUPROR ?SOY, The Solae Company,
St. Louis, MO, formerly DuPont Technologies), providing
40 g of soy protein and 80 mg of isoflavones (aglycone units)
daily. Control group participants were under medical treat-
ment by their personal physician.
Baseline and one-yr nutrient intake data were collected
using 3-day food diaries. Dietary intake data were analyzed
using Nutrition Data System for Research (NDS-R) software
versions4.01 29and4.02 30(NutritionCoordinatingCenter
(NCC), University of Minnesota, Minneapolis, MN). Final
NDS-R time-related database updates analytic data while
maintaining nutrient profiles true to the version used for data
collection. Isoflavone values specific for the soy protein sup-
plement used in the intervention were provided by the man-
ufacturer and substituted for the isoflavone value estimated
Serum IGF-I, IGFBP-1, IGFBP-2, IGFBP-3, and insulin
were measured in duplicate using commercial ELISA kits
(Diagnostic Systems Laboratories, Webster, TX). All assays
were performed in a blinded manner, and control samples
provided by the manufacturer were included in each run.
IGF-1 was separated from its binding proteins in serum prior
to measurement to obtain total IGF-1. The IGF-1: IGFBP-
3 molar ratio, an indicator of bioactive IGF-1, was calcu-
lated using the following conversion factors: 1 ng/ml IGF-
1 = 0.130 nm IGF-1, and 1 ng/ml IGFBP-3 = 0.036 nm
IGFBP-3. Fasting insulin was measured using the DSL-10-
1600 ACTIVE Insulin ELISA Kit (Diagnostic Systems Lab-
oratories, Inc., Webster, TX). Serum was also used to study
serum-stimulated growth and apoptosis of androgen depen-
dent LNCaP cell line (27). Serum PSA was measured at
Memorial Sloan-Kettering Cancer Center prospectively by a
heterogeneous sandwich magnetic separation assay with the
Immuno 1 System. Testosterone was measured by a com-
petitive immunoassay with an ImmuliteR ?automated an-
alyzer. All serum markers were measured at baseline and
Independent samples t-tests tested for equivalency be-
tween experimental groups at baseline. Experimental group
differences in changes from baseline to one year in soy
isoflavone and protein intakes (total, animal, and vegetable),
IGF-1, IGF binding proteins, the IGF-1: IGFBP-3 molar ra-
tio, and fasting insulin were analyzed using analysis of vari-
ance for repeated measures, with experimental group as a
between subjects factor and time as a repeated factor. Cross-
sectional relationships at baseline (entire sample) between
ing the IGF axis (IGF-1, IGFBP-1, IGFBP-2, and IGFBP-3,
cell growth and apoptosis, PSA and testosterone were ana-
lyzed using Pearson correlations (due to low intake at base-
line, soy isoflavones were not included in the analyses). We
also repeated these analyses including soy isoflavones intake
on data obtained at 1 yr. Furthermore, Pearson correlations
were used to evaluate the relationship between changes in
these variables over 1 yr for the intervention group par-
ticipants. Additionally, using a median split on protein in-
take, outcomes for the intervention group patients with high
36Nutrition and Cancer 2007
Table 1. Participant Characteristics at Baseline1
(n = 44)
(n = 49)
Total Protein (g)
Animal Protein (g)
Vegetable Protein (g)
Total Isoflavones (mg)
Isoflavones from diet (mg)
Isoflavones from Suppl. (mg)
IGF-1 : IGFBP-3 (molar ratio)
Fasting Insulin (µIU/ml)
81 ± 26
40 ± 23
41 ± 18
19 ± 29
17 ± 26
1 ± 5
79 ± 23
40 ± 22
38 ± 22
16 ± 37
15 ± 36
1 ± 4
168 ± 64
30 ± 22
459 ± 246
1792 ± 365
0.35 ± 0.17
153 ± 62
26 ± 18
498 ± 215
1699 ± 386
0.32 ± 0.12
6.8 ± 57.6 ± 70.50
insulin-like growth factor binding protein; µIU, micro-International units;
FBS, fetal bovine serum; PSA, prostate specific antigen. 2: n = 42 for
interventiongroupandN =43forcontrolgroup.3:n =40forintervention
group and N = 40 for control group. 4: n = 39 for intervention group and
N = 40 for control group.
protein intake at 1 yr were compared to outcomes of those
with low protein intake using independent samples t-tests.
have been described in detail previously (27,29); baseline 3-
inTable 1.No significant differences were observed between
the two groups at baseline in diet, IGF axis or fasting insulin
variables (one patient’s fasting insulin level of 221 µIU/ml,
Table 2. Correlations of Age, Weight, Exercise, and Protein Intake with IGF Axis, Insulin Levels, and Prostate Cancer
Markers at Baseline1,2
VariableAge (y) Weight (kg)Exercise (hrs) Total Protein (g)Animal Protein (g) Vegetable Protein (g)
IGF-1:IGFBP-3 (molar ratio)
Fasting Insulin (µIU/ml)
Prostate Cancer Markers
LNCaP apoptosis (% FBS)
LNCaP growth (% FBS)
1: Abbreviations are as follows: IGF-1, insulin-like growth factor 1; IGFBP, insulin-like growth factor binding protein; µIU, micro-International units; FBS,
fetal bovine serum; PSA, prostate specific antigen. 2: For protein data, n = 77 for correlations with IGF axis, LNCaP apoptosis and growth, n = 76 with
insulin, n = 85 with PSA, n = 78 with testosterone. For age, weight, and exercise data, n = 80 for correlations with IGF axis, LNCaP apoptosis and growth,
n = 79 with insulin, n = 93 with PSA, n = 81 with testosterone. Soy isoflavone data were not included due to low variability of soy isoflavone intake at
baseline. * P < 0.05. ** P < 0.01. *** P < 0.001.
far greater than 3 standard deviations from mean, was not
Baseline correlations of protein intake, exercise, weight
and age, to prostate cancer markers in the entire sample are
seen in Table 2. Higher age was associated with decreased
IGF-1, IGF-1: IGFBP-3 molar ratio, and lower LNCaP cell
growth. Higher levels of exercise were associated with lower
IGF-I and IGF-1: IGFBP-3 molar ratio. Higher weight was
associated with lower levels of IGFBP-1 and IGFBP-2 and
with higher levels of fasting insulin. Higher levels of di-
etary protein from both animal and vegetable sources were
tion,higher consumption of totalproteinwas associated with
lower testosterone levels, and higher consumption of animal
protein was associated with higher fasting insulin and lower
IGFBP-1. Higher levels of vegetable protein were also as-
sociated with increased LNCaP cell growth. In addition, at
baseline, fasting insulin was negatively correlated to IGFBP-
1 (r = –.418, P < 0.001) and IGFBP-2 (r = –.332, P <
0.001), but showed no relationship to IGFBP-3.
Experimental group differences in changes from baseline
yses involving dietary variables are based on 74 participants
(Table 3). Follow-up analyses involving IGF axis and insulin
At 1 yr, intervention patients reported higher intakes of total
and vegetable protein, lower intakes of animal protein and
higher intakes of soy isoflavones, both from foods and the
no changes in these variables at 1 yr.
to 1 yr were observed regardless of experimental grouping.
Intervention patients, but not control patients, demonstrated
increases in IGFBP-1 at 1 yr. There were no statistically
Vol. 58, No. 1 37
Table 3. Protein and Isoflavone Intake by Experimental Group and Time Period, N = 74
P-value Time × Group
Total protein (g/day)
Animal protein (g/day)
Vegetable protein (g/day)
Total isoflavones (mg/day)
Isoflavones from diet (mg/day)
Isoflavones from suppl.(mg/day)
0.001 0.001 0.001
1: Mean ± standard deviation.
significant differences between experimental and control pa-
tients for IGFBP-2 or the IGF-1: IGFBP-3 molar ratio. Al-
though there was a significant group by time interaction in
fasting insulin, post-hoc adjustments did not reveal any sig-
nificant differences between the two groups during the 1-yr
For the intervention group, there were sufficient changes
in isoflavone and protein intake from baseline to one year to
examine the relationship between these changes and changes
in IGF axis parameters, insulin, and prostate cancer markers
protein consumption were highly correlated with changes in
soy isoflavones (r = .70, P < 0.001). Increases in vegetable
Table 4. IGF Axis and Insulin Levels by Experimental Group and Time period1,2
P-value Time × Group
IGF-1: IGFBP-3 (molar ratio)
Fasting Insulin (µIU/ml)
168 ± 643
153 ± 62
199 ± 83
170 ± 85
0.02 0.01 0.001
459 ± 246
498 ± 215
501 ± 255
526 ± 228
1792 ± 365
1699 ± 386
1894 ± 378
1825 ± 422
0.35 ± 0.17
0.32 ± 0.12
0.38 ± 0.17
0.33 ± 0.13
1: Abbreviations are as follows: IGF-1, insulin-like growth factor 1; IGFBP, insulin-like growth factor binding protein; IU, international units. 2: n = 80 for
IGF axis variables (40 in the intervention group, 40 in the control group); n = 79 for fasting insulin (39 in the intervention group, 40 in the control group). 3:
Mean ± standard deviation.
protein intake were associated with increases in IGFBP-1
(P < 0.05). No other correlations were significant. Correla-
year (not shown) revealed that higher intakes of total protein
were associated with higher levels of IGFBP-1 (P < 0.01)
higher consumption of vegetable protein, in contrast with
baseline correlations, was associated with higher levels of
IGFBP-1 and with lower levels of LNCaP cell growth (P <
0.001 for both). Conversely, higher intakes of animal protein
were associated with lower levels of IGFBP-1 (P < 0.001),
higher levels of LNCaP cell growth (P < 0.001), higher
PSA (P < 0.05) and higher fasting insulin (P < 0.01). In
38 Nutrition and Cancer 2007
addition, higher consumption of vegetable protein was as-
sociated with lower levels of fasting insulin (P < 0.01).
Correlations between soy isoflavone intakes and biomedi-
cal variables at one year (not shown) revealed that higher
intakes of soy isoflavones (from both diet and supplement)
were associated with higher levels of IGFBP-1 (P < 0.01),
lower levels of LNCaP cell growth (P < 0.05) and lower
levels of fasting insulin (P < 0.05).
changes in the IGF axis (not shown) revealed a negative
relationship of insulin to IGFBP-1 (r = –.270, P < 0.05)
and a positive relationship of insulin to IGFBP-3 (r = .313,
P < 0.01).
Comparisons of experimental patients whose protein in-
take at 1 yr was above the median (117 g/day) to those below
the median did not reveal significant differences on any vari-
The results of this study indicate that high dietary intakes
of protein and soy isoflavones, as a measure of soy protein
intake, in the context of comprehensive lifestyle changes,
do not appear to be responsible for the increase in IGF-1
observed after one year in prostate care patients in both the
intervention and control groups. Furthermore, our analyses
indicate that, in these patients, increased consumption of
vegetable protein was associated with increased levels of
Analyses of baseline data show that this population re-
ported a similar intake of protein (16% of energy) as typi-
cally consumed by Americans in the same age group (31).
Soy isoflavone intake was minimal, as usually seen in West-
ern diets (32). As expected, higher levels of exercise were
IGFBP-3 molar ratio. We also found that, at baseline, dietary
intake of protein from all sources was positively correlated
with the IGF-1: IGFBP-3 molar ratio, in agreement with
higher levels of vegetable protein to be associated with in-
when considering changes in these variables from baseline
to 1 yr and was reversed in analyses on 1-yr data. That is, an
inverse association between vegetable protein consumption
needs to be further investigated.
As was expected with the adoption of a vegan diet sup-
plemented with soy protein, intervention group participants
protein and decreased intake of animal protein after 1 yr. As
ings of soy products and to include a 58-g serving of a soy
protein supplement daily, soy isoflavone intake also signif-
icantly increased, equally from dietary sources and the soy
protein supplement. In fact, the increases in protein intake
from baseline to 1 yr in the experimental group were highly
correlated with increases in soy isoflavones. The daily soy
isoflavone levels achieved by this sample (133 mg/day) sig-
nificantly exceed the average daily intake in Japanese, which
has been reported to range from 30 to 50 mg/day (33).
Interventions similar to the one used in this study, in-
cluding a diet providing <10% energy from fat and moder-
ate amounts of protein (15–20% energy) together with ∼60
been shown to decrease IGF-1 levels in overweight men in
the short-term (11 days) and long-term (14 yr) (5,6). In con-
trast, in the present study, IGF-1 significantly increased in
both experimental and control groups, while remaining, on
average, within the normal range (34). It is possible that the
tal group, from an average of 80 g/day (16% of total energy)
to 115 g/day, (20% of total energy), may have mitigated the
potential IGF-1 lowering effect of the very-low fat diet and
increased exercise. The protein level consumed by experi-
mental participants was noticeably higher than the dietary
reference intake (DRI) for adult men, which for this group
would amount to 60–72 g protein/day (based on 0.8–0.96
g protein/kg body weight for omnivores-vegans) (35,36).
Nonetheless, the somewhat smaller, but statistically signifi-
cant, rise in IGF-1 in the control group remains unexplained.
Although patients with prostate cancer show levels of IGF-1
that are approximately 8% higher than men without prostate
cancer (37), we are not aware of any studies describing the
progression of circulating IGF-1 over time in patients with
early-stage prostate cancer. In addition, although an associ-
ation between IGF-1 and risk of prostate cancer has been
observed, the predictive value of plasma concentrations of
IGF-1 and its binding proteins in the prognosis of prostate
cancer is still being debated (4,37–40).
The level of dietary protein achieved by experimental
patients at 1 yr (115 g/day) is similar to the highest quintile
of protein intake in the Health Professionals Follow-Up
study (107 g/day) (9). Although our dietary analysis did not
allow us to distinguish soy protein from vegetable protein, it
is reasonable to conclude, from the high correlation between
changes in protein and soy isoflavone intake from baseline
to 1 yr, that most of the protein increase that incurred
in the intervention group was due to the addition of soy
protein, both from soy products and from the supplement.
Interestingly, Allen (12) observed that soy protein intake
was positively correlated to IGF-1 in vegans, whereas plant
protein from sources other than soy was inversely correlated
to IGF-1, confirming previous findings that protein high
in essential amino acids is an important determinant of
circulating IGF-1 (15). Indeed, the food groups associated
with higher levels of IGF-1 are milk, dairy products, fish,
poultry, and red meat in omnivorous populations (9–11)
and soymilk in vegans (12), all sources of protein high
in essential amino acids. Observational studies have also
indicated that intake of soy is associated with higher IGF-1
levels in men, although not in women (22), possibly because
of gender differences in endogenous estrogen levels (32). It
Vol. 58, No. 139
IGF-1 (23,24), randomized controlled trials using both an
intact soy protein and soy protein deprived of isoflavones
have demonstrated that the isoflavone content of soy is not
responsible for IGF-1-raising effect of soy protein (24,25).
The finding of an increase in IGFBP-3 is in disagree-
ment with investigations showing that low-fat diets (with or
without exercise), and/or soy protein supplementation do not
result in any significant short-term or long-term changes in
IGFBP-3 (5,26). By contrast, Gann et al. showed a small
but statistically significant reduction in IGFBP-3, and an in-
crease in the IGF-1: IGFBP-3 molar ratio, in women follow-
ing a low-fat diet (< 20% energy from fat) supplemented
with 40 g of a daily soy protein supplement (25). Similarly,
supplementation with 40 g of soy protein for 12 mo in a pre-
in IGFBP-3, and a significant increase in serum IGF-1 and
the IGF-1: IGFBP-3 molar ratio compared to baseline (24).
in the intervention group compared to controls over the
study period. Although the change in IGFBP-1 over time
was modest (33%), this finding is consistent with recent data
on the effect of a very low-fat diet and exercise program in
men (5,6). These investigations showed that a diet very low
in fat and moderate in protein, together with daily aerobic
exercise, resulted in significant increases in IGFBP-1 and
significant reductions in IGF-1 compared to baseline levels
(5) or to a control group (6). In addition, serum-stimulated
LNCaP cell growth was reduced and apoptosis in LNCaP
cells incubated with post-intervention serum was increased
compared to baseline levels and to controls. Interestingly,
serum IGFBP-1 levels showed an inverse relationship,
whereas IGF-1 showed a positive relationship, with LNCaP
cell growth (5). In the PCLT, we reported a significant
decrease in serum-stimulated LNCaP cell growth and a
non significant increase in LNCaP cell apoptosis in the
intervention group compared to controls (27).
Unlike the Pritikin program studies (5,6), we did not find
a significant reduction in fasting insulin in experimental pa-
tients over the study period, possibly due to lack of statistical
power. However, there was a trend towards lower fasting
insulin levels in the experimental group compared to the
controls. This is encouraging, as fasting insulin has been
associated with increased risk of prostate cancer (41). As ex-
pected, fasting insulin was inversely correlated to IGFBP-1
at baseline and changes in fasting insulin were negatively
associated to changes in IGFBP-1 at 1 yr (42).
The examination of the relationship between changes in
soy isoflavone and protein intakes from baseline to one year
and changes in prostate cancer markers revealed that higher
intakes of vegetable protein were associated with increases
in IGFBP-1. This finding is consistent with observations that
vegans have higher IGFBP-1 than meat-eaters or vegetari-
ans (12,43). In addition, the correlations performed at one
year confirmed the positive association between IGFBP-1
and vegetable protein, and also showed a positive associa-
tion with soy isoflavones. Furthermore, both vegetable pro-
isoflavones may be beneficial against tumor growth (18–20).
Our study is limited by the post-hoc design, which did
not allow us to analyze the relationship of dietary protein
and isoflavones alone to the IGF axis. Due to the presence of
multiple components in this intervention, we cannot rule out
the influence of changes other than protein and isoflavone in-
take to the IGF axis. Nonetheless, the observed group differ-
ences in protein and isoflavone intakes are substantial (44%
of patients with early-stage prostate cancer who elected ac-
to be generalized to all men with prostate cancer. Another
limitation is the lack of measurement of plasma isoflavones
as evidence of compliance with the diet protocol, and of
plasma steroid hormones other than testosterone that could
have provided additional explanation for our results.
In summary, comprehensive lifestyle changes including a
patients randomized to the experimental group compared
to controls. Contrary to our expectations, we did not find
that either dietary protein or soy isoflavones, as an indicator
of soy protein consumption, were significantly related to
IGF-1. Nonetheless, given the recent findings that protein
rich in essential amino acids (animal and soy protein) is
associated with increases in IGF-1, it may be prudent for
men with early stage prostate cancer consuming a vegan diet
already adequate in dietary protein, not to exceed the protein
recommendations set by the Institute of Medicine (35).
Acknowledgments and Notes
Supported by the Department of Defense Uniformed Services Uni-
versity Grant MDA905-99-1-0003 via the Henry M. Jackson Foundation
Grant 600-06971000-236 Department of the Army (U.S. Army Medical
Research Acquisition Activity W81XWH-05-1-0375-P0001), the Depart-
ment of Health and Human Services (Health Resources and Services Ad-
ministration #4 C76HF00803-01-01), The Prostate Cancer Foundation, Na-
tional Institutes of Health 5P50CA089520-02 University of California–San
Francisco Prostate Cancer Specialized Program of Research Excellence, the
Safeway Foundation, and the Walton Family Foundation. Representatives
Nancy Pelosi and John Murtha, and Senators Arlen Specter and Ted Stevens
provided support. Insulin and IGF proteins analyses were done by Dr. Rusty
Nicar at Diagnostic System Laboratory. We thank Colleen Kemp for her
Technologies, for providing the soy protein powdered beverage. Address
correspondence to Gerdi Weidner, Preventive Medicine Research Institute,
900 Bridgeway, Sausalito, CA 94965. E-mail: email@example.com.
Submitted 18 May 2006; accepted in final form 9 January 2007.
1. Djavan B, Waldert M, Seitz C, and Marberger M: Insulin-like growth
factors and prostate cancer. World J Urol 19, 225–233, 2001.
40 Nutrition and Cancer 2007
2. Yu H and Rohan T: Role of the insulin-like growth factor family in
cancer development and progression. J Natl Cancer Inst 92, 1472–
3. Renehan AG, Zwahlen M, Minder C, O’Dwyer ST, Shalet SM, et al.:
Insulin-like growth factor (IGF)-I, IGF binding protein-3, and can-
cer risk: systematic review and meta-regression analysis. Lancet 363,
4. Chan JM, Stampfer MJ, Ma J, Gann P, Gaziano JM, et al.: Insulin-
like growth factor-I (IGF-I) and IGF binding protein-3 as predictors
of advanced-stage prostate cancer. J Natl Cancer Inst 94, 1099–1106,
5. Ngo TH, Barnard RJ, Tymchuk CN, Cohen P, and Aronson WJ: Effect
of diet and exercise on serum insulin, IGF-I, and IGFBP-1 levels and
growthofLNCaP cellsin vitro (United States).CancerCausesControl
13, 929–935, 2002.
6. Barnard RJ, Ngo TH, Leung PS, Aronson WJ, and Golding LA: A low-
prostate tumor cell growth in vitro. Prostate 56, 201–206, 2003.
7. Chan JM, Stampfer MJ, Giovannucci E, Gann PH, Ma J, et al.: Plasma
Science 279, 563–566, 1998.
8. Chokkalingam AP, Pollak M, Fillmore CM, Gao YT, Stanczyk FZ, et
al.: Insulin-like growth factors and prostate cancer: a population-based
case-control study in China. Cancer Epidemiol Biomarkers Prev 10,
predictors of insulin-like growth factor I and their relationships to
cancer in men. Cancer Epidemiol Biomarkers Prev 12, 84–89, 2003.
10. Holmes MD, Pollak MN, Willett WC, and Hankinson SE: Dietary
factor binding protein 3 concentrations. Cancer Epidemiol Biomarkers
Prev 11, 852–861, 2002.
11. Larsson SC, Wolk K, Brismar K, and Wolk A: Association of diet with
serum insulin-like growth factor I in middle-aged and elderly men. Am
J Clin Nutr 81, 1163–1167, 2005.
12. Allen NE, Appleby PN, Davey GK, Kaaks R, Rinaldi S, et al.: The
associations of diet with serum insulin-like growth factor I and its main
binding proteins in 292 women meat-eaters, vegetarians, and vegans.
Cancer Epidemiol Biomarkers Prev 11, 1441–1448, 2002.
of the insulin-like growth factors. Endocr Rev 15, 80–101, 1994.
14. Smith WJ, Underwood LE, and Clemmons DR: Effects of caloric or
protein restriction on insulin-like growth factor-I (IGF-I) and IGF-
binding proteins in children and adults. J Clin Endocrinol Metab 80,
amino acids augment the somatomedin-C/insulin-like growth factor I
response to refeeding after fasting. Metabolism 34, 391–395, 1985.
16. Phillips LS, Harp JB, Goldstein S, Klein J, and Pao CI: Regulation and
action of insulin-like growth factors at the cellular level. Proc Nutr Soc
49, 451–458, 1990.
17. Sim HG and Cheng CW: Changing demography of prostate cancer in
Asia. Eur J Cancer 41, 834–845, 2005.
18. Zhou JR, Gugger ET, Tanaka T, Guo Y, Blackburn GL, et al.: Soybean
cinoma and tumor angiogenesis in mice. J Nutr 129, 1628–1635, 1999.
19. Zhou JR, Yu L, Zhong Y, Nassr RL, Franke AA, et al.: Inhibition
of orthotopic growth and metastasis of androgen-sensitive human
prostate tumors in mice by bioactive soybean components. Prostate
53, 143–153, 2002.
20. Lamartiniere CA, Cotroneo MS, Fritz WA, Wang J, Mentor-Marcel
R, et al.: Genistein chemoprevention: timing and mechanisms of
action in murine mammary and prostate. J Nutr 132, 552S–558S,
21. Jarred RA, Keikha M, Dowling C, McPherson SJ, Clare AM, et
al.: Induction of apoptosis in low to moderate-grade human prostate
Biomarkers Prev 11, 1689–1696, 2002.
nants of circulating insulin-like growth factor I and insulin-like growth
factor binding protein 3 concentrations in a cohort of Singapore men
and women. Cancer Epidemiol Biomarkers Prev 12, 739–746, 2003.
23. Khalil DA, Lucas EA, Juma S, Smith BJ, Payton ME, et al.: Soy
protein supplementation increases serum insulin-like growth factor-I
in young and old men but does not affect markers of bone metabolism.
J Nutr 132, 2605–2608, 2002.
24. Adams KF, Newton KM, Chen C, Emerson SS, Potter JD, et al.: Soy
isoflavones do not modulate circulating insulin-like growth factor
concentrations in an older population in an intervention trial. J Nutr
133, 1316–1319, 2003.
25. Gann PH, Kazer R, Chatterton R, Gapstur S, Thedford K, et al.:
Sequential, randomized trial of a low-fat, high-fiber diet and soy
supplementation: effects on circulating IGF-I and its binding proteins
in premenopausal women. Int J Cancer 116, 297–303, 2005.
naive prostate cancer patients. Clin Cancer Res 9, 3282–3287, 2003.
27. Ornish D, Weidner G, Fair WR, Marlin R, Pettengill EB, et al.:
Intensive lifestyle changes may affect the progression of prostate
cancer. J Urol 174, 1065–1069; discussion 1069–1070, 2005.
28. Dunn-Emke SR, Weidner G, Pettengill EB, Marlin RO, Chi C, et al.:
Nutrient adequacy of a very low-fat vegan diet. J Am Diet Assoc 105,
29. Daubenmier JJ, Weidner G, Marlin R, Crutchfield L, Dunn-Emke S,
et al.: Lifestyle and health-related quality of life of men with prostate
cancer managed with active surveillance. Urology 67, 125–130, 2006.
30. Ornish DM, Lee KL, Fair WR, Pettengill EB, and Carroll PR: Dietary
trial in prostate cancer: Early experience and implications for clinical
trial design. Urology 57, 200–201, 2001.
31. Wright JD, Wang CY, Kennedy-Stephenson J, and Ervin RB: Dietary
intake of ten key nutrients for public health, United States: 1999–2000.
Adv Data 1–4, 2003.
32. Vrieling A, Voskuil DW, Bueno de Mesquita HB, Kaaks R, van Noord
PA, et al.: Dietary determinants of circulating insulin-like growth
factor (IGF)-I and IGF binding proteins 1, -2, and -3 in women in the
Netherlands. Cancer Causes Control 15, 787–796, 2004.
33. Wakai K, Egami I, Kato K, Kawamura T, Tamakoshi A, et al.: Dietary
intake and sources of isoflavones among Japanese. Nutr Cancer 33,
34. Yu H, Mistry J, Nicar MJ, Khosravi MJ, Diamandis A, et al.: Insulin-
like growth factors (IGF-I, free IGF-I and IGF-II) and insulin-like
growth factor binding proteins (IGFBP-2, IGFBP-3, IGFBP-6, and
ALS) in blood circulation. J Clin Lab Anal 13, 166–172, 1999.
35. Food and Nutrition Board Institute of Medicine: Dietary Reference
Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol,
Protein, and Amino Acids (Macronutrients). Washington, DC: The
National Academy Press, 2002.
36. Position of the American Dietetic Association and Dietitians of
Canada: Vegetarian diets. J Am Diet Assoc 103, 748–765, 2003.
37. Cohen P: Serum insulin-like growth factor-I levels and prostate cancer
risk–interpreting the evidence. J Natl Cancer Inst 90, 876–879, 1998.
38. Harman SM, Metter EJ, Blackman MR, Landis PK, and Carter
HB: Serum levels of insulin-like growth factor I (IGF-I), IGF-II,
IGF-binding protein-3, and prostate-specific antigen as predictors of
clinical prostate cancer. J Clin Endocrinol Metab 85, 4258–4265,
39. Yu H, Nicar MR, Shi R, Berkel HJ, Nam R, et al.: Levels of insulin-like
growth factor I (IGF-I) and IGF binding proteins 2 and 3 in serial
postoperative serum samples and risk of prostate cancer recurrence.
Urology 57, 471–475, 2001.
Circulating free insulin-like growth factor (IGF)-I, total IGF-I, and
IGF binding protein-3 levels do not predict the future risk to develop
prostate cancer: results of a case-control study involving 201 patients
within a population-based screening with a 4-year interval. J Clin
Endocrinol Metab 89, 4391–4396, 2004.
Vol. 58, No. 1 41
41. Hsing AW, Chua S, Jr., Gao YT, Gentzschein E, Chang L, et al.:
Prostate cancer risk and serum levels of insulin and leptin: a
population-based study. J Natl Cancer Inst 93, 783–789, 2001.
42. Suikkari AM, Koivisto VA, Rutanen EM, Yki-Jarvinen H, Karonen
SL, et al.: Insulin regulatesthe serum levels of low molecular weight
insulin-like growth factor-binding protein. J Clin Endocrinol Metab
66, 266–272, 1988.
43. Allen NE, Appleby PN, Davey GK, and Key TJ: Hormones and diet:
low insulin-like growth factor-I but normal bioavailable androgens in
vegan men. Br J Cancer 83, 95–97, 2000.
42Nutrition and Cancer 2007