Internal Medicine

Published on January 2017 | Categories: Documents | Downloads: 36 | Comments: 0 | Views: 423
of 120
Download PDF   Embed   Report

Comments

Content


Benign Prostatic Hyperplasia: Role Of Metabolic Abnormalities And Association
With Prostate Cancer Risk
Jeannette Marie Kisser
A dissertation submitted in partial fulfillment of the
requirements for the degree of
Doctor of Philosophy
University of Washington
2010
Program Authorized to Offer Degree:
School of Public Health-Department of Epidemiology
UMI Number: 3406118
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMT
Dissertation Publishing
UMI 3406118
Copyright 2010 by ProQuest LLC.
All rights reserved. This edition of the work is protected against
unauthorized copying under Title 17, United States Code.
ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
University of Washington
Graduate School
This is to certify that I have examined this copy of a doctoral dissertation by
Jeannette Marie Kisser
and have found that it is complete and satisfactory in all respects, and that any and all
revisions required by the final examining committee have been made.
Chair of the Supervisory Committee:
Reading Committee:
Alan R. Kristal
Emily White
Catherine Tangen
^ .
Date: 'fl^VWiUAi Vj
]
ZQ\Q
In presenting this dissertation in partial fulfillment of the requirements for the doctoral degree at the
University of Washington, I agree that the Library shall make its copies freely available for inspection. I
further agree that extensive copying of this dissertation is allowable only for scholarly purposes,
consistent with "fair use" as prescribed in the U.S. Copyright Law. Requests for copying or reproduction
of this dissertation may be referred to ProQuest Information and Learning, 300 North Zeeb Road, Ann
Arbor, Ml 84106-1346,1-800-521-0600 to whom the author has granted "the right to reproduce and sell
(a) copies of the manuscript in microform and/or (b) printed copies of the manuscript made from
microform.
Signature:
0
Date:
University of Washington
Abstract
Benign Prostatic Hyperplasia: Role Of Metabolic Abnormalities And Association
With Prostate Cancer Risk
Jeannette Marie Kisser
Chair of the Supervisory Committee:
Professor Alan R. Kristal
Department of Epidemiology
This dissertation examines the role of obesity related metabolic abnormalities in the
development of symptomatic benign prostatic hyperplasia (BPH), and investigates the
association between symptomatic BPH and prostate cancer. Data are from the Prostate Cancer
Prevention Trial (PCPT), a 7-year randomized, double-blinded, placebo-controlled trial of the
drug finasteride (Proscar®) for the primary prevention of prostate cancer. Analyses are
restricted to the approximate 9,000 men in the placebo control group. This dissertation begins
by examining whether blood concentrations of two adipokines, adiponectin and leptin, and C-
peptide, an indirect measure of insulin secretion and insulin resistance affect the risk of
symptomatic BPH. This analysis shows that high adiponectin concentrations were associated
with a reduced risk of symptomatic BPH; however, only for men who were moderately/very
active. The second analysis examines whether blood concentrations of pro-
inflammatory cytokines known to be elevated in obesity are associated with risk of
symptomatic BPH. This analysis shows that low soluble tumor necrosis factor receptor
2 and high interleukin 6 concentrations increased BPH risk; however, only among men
aged < 65 years. Together, these results suggest that obesity related metabolic abnormalities
in glucose uptake and insulin sensitivity and systemic inflammation increase the risk of
symptomatic BPH; however, these abnormalities explain only a moderate amount of the
relationship between obesity and BPH. Finally, I explore the association between symptomatic
BPH and risk of prostate cancer in a study population where diagnostic surveillance of
both BPH and prostate cancer was rigorous and complete. In this analysis there were no
associations of prevalent BPH or prevalent plus incident BPH with prostate cancer risk.
This lack of association was also consistent across subgroups defined by type of BPH
defining event (treatment, symptoms or physician diagnosis), prompt or reason of
prostate cancer diagnosis ('For-cause' or 'Not-for-cause') and prostate cancer grade
(Gleason score 2-7(3+4) vs. 7(4+3)-10). This study provides the strongest evidence to
date that BPH does not increase the risk of prostate cancer.
Table of Contents
Page
List of Tables ii
Acknowledgements v
Dedication vi
Introduction 1
Chapter 1. Serum Adiponectin, C-Peptide and Leptin and Risk of
Symptomatic Benign Prostatic Hyperplasia 19
Chapter 2. Biomarkers of Systemic Inflammation and Risk of
Incident, Symptomatic Benign Prostatic Hyperplasia 42
Chapter 3. Benign Prostatic Hyperplasia and Risk of Prostate Cancer 72
Reference List 100
I
List of Tables
Table Number Page
Table 1.1. Demographic and lifestyle characteristics of the study
population 33
Table 1.2. Serum concentration of adipokines and C-peptide for cases
and controls 34
Table 1.3. Association of adipokines and C-peptide on BPH risk 35
Table 1.4. Associations of Adiponectin, C-peptide and Leptin (quartiles)
with risk of BPH categorized as clinical or symptomatic progression BPH 36
Table 1.5. Association of adipokines and C-peptide on BPH risk, excluding
men with prostate cancer 37
Table 1.6. Associations of Adiponectin, C-peptide and Leptin (quartiles)
with risk of BPH categorized as BPH cases captured in years 1-3 or 4-7 38
Table 1.7. Association of adipokines and C-peptide with BPH risk,
stratified by physical activity, alcohol consumption and BMI 39
Table 1.8. Main effects of BMI on BPH risk, adjusted for adipokines and
C-peptide 41
Table 2.1. Demographic and lifestyle characteristics of the study
population 58
Table 2.2. Serum concentrations of cytokines for cases and
controls 59
Table 2.3. Main effects of CRP and cytokines on BPH risk 60
n
List of Tables (continued)
Table Number Page
Table 2.4. Main effects of CRP and cytokines on BPH risk, excluding
men with prostate cancer 61
Table 2.5. Main effects of CRP and cytokines on BPH risk, excluding
men who died during the PCPT 62
Table 2.6. Associations of CRP and cytokines with BPH risk, stratified
by body mass index 63
Table 2.7. Associations of CRP and cytokines with BPH risk, stratified
by smoking status 64
Table 2.8. Associations of CRP and cytokines with BPH risk, stratified
by physical activity 65
Table 2.9. Associations of CRP and cytokines with BPH risk, stratified
by age 66
Table 2.10. Associations of CRP and cytokines with BPH risk, stratified
by type of BPH defining event 67
Table 2.11. Associations of CRP and cytokines with BPH risk, stratified
by time from baseline to BPH event 68
Table 2.12. Associations of CRP, sTNF-RII and IL-6 with BPH risk,
stratified by type of LUTS 69
Table 2.13. Associations of CRP, sTNF-RII and IL-6 with BPH risk among
young men only, stratified by type of LUTS 70
in
List of Tables (continued)
Table Number Page
Table 2.14. Association between obesity and BPH risk adjusted for one
or more biomarkers of systemic inflammation 71
Table 3.1. Distribution of demographic and lifestyle characteristics among
the total study population 88
Table 3.2. Percent of men with prevalent BPH by demographic and
lifestyle characteristics 90
Table 3.3. Percent of men with prevalent plus incident BPH by
demographic and lifestyle characteristics 91
Table 3.4. Percent of men with prostate cancer by demographic and
lifestyle characteristics 93
Table 3.5. Associations of prevalent Benign Prostatic Hyperplasia with
prostate cancer risk 95
Table 3.6. Associations of prevalent plus incident Benign Prostatic
Hyperplasia with prostate cancer risk in the Prostate Cancer
Prevention Trial 96
Table 3.7. Associations of prevalent Benign Prostatic Hyperplasia with
prostate cancer risk; High grade cancers defined as Gleason grade 8+ 97
Table 3.8. Associations for prevalent plus incident Benign Prostatic Hyperplasia
with prostate cancer risk; High grade cancers defined as Gleason grade 8+ 98
IV
Acknowledgements
Financial support for this project was provided by grant 5R01DK063303-02, the
Cohort Study of Risks for Benign Prostatic Hyperplasia, and training grant number
R25 CA94880, Cancer Prevention Training: Nutrition, Exercise, and Genetics, both
from the National Institutes of Health.
Thank you to the chair of my committee Dr. Alan Kristal for the general academic and
professional support during this work, and for spending many hours helping me to
become a better writer.
Thank you to the members of my committee Dr. Marian Neuhouser, Dr. Catherine
Tangen, Dr. Emily White and Dr. Daniel Lin for providing constructive feedback and
encouragement.
Thank you to Ms. Katherine Arnold helping fulfill my multiple data requests and for
explaining the intricacies of the PCPT dataset.
Thank you to Ms. MJ Welling for providing editorial support for this work.
v
Dedication
This dissertation is dedicated to my parents, Joseph and Marlene Schenk, for being
supportive of every career and educational decision I made, no matter how strange, and
to my husband, Scott Kisser, for all the extra cooking, dishes, dog walks, patience, and
overall love and support he provided throughout this process
VI
7
Introduction
Benign prostatic hyperplasia (BPH) is one of the most common medical
conditions in middle-aged and older males. Estimates of BPH prevalence are
imprecise due to a lack of uniform diagnostic criteria[l], but range from 40-50% at 50
years to as high as 80% for men in their 70s.[2-4]
The prostate is a complex and highly heterogenous tissue comprised of four
distinct regions: a non-glandular fibromuscular stroma and three glandular zones (the
peripheral, central and transition zones. BPH is characterized by the formation of large
nonmalignant nodules, followed by glandular hyperplasia in the periuretheral or
transition zone of the prostate. [5] As BPH progresses, the gland enlarges and
compresses the urethra, which disrupts the flow of urine and contributes to the lower
urinary tract symptoms (LUTS) characteristic of the disease. [6] BPH is rarely a life-
threatening condition; however, the symptoms associated with it can be very
bothersome can interfere with usual daily activities and have a marked effect on quality
of life. [7-9] Left untreated, BPH may progress in severity and result in urological
complications such as recurrent bladder infections, bladder calculi, or urinary retention,
which often require medical treatment or prostate surgery.[ 10] Some researchers have
even hypothesized that BPH may be related to risk of prostate cancer. [11]
Obesity and BPH
The pathogenesis of BPH remains poorly understood; however, there is
consistent evidence from several epidemiologic studies that establishes obesity as a risk
8
factor for BPH. Of the 7 studies that have examined the risk of surgical and/or medical
treatment for BPH[12-15,16)(Seitter, 1992 #1501,17], more than half found a
significant increase in risk that was specific to abdominal obesity.[12-15] The largest
and best designed of these studies include the Prostate Cancer Prevention Trial and the
Health Professionals Follow-up Study cohorts, both of which reported a significantly
increased risk of symptomatic BPH (defined by self-reported lower urinary tract
symptoms) for abdominal obesity.[12,13] Several cross-sectional studies have also
examined the association between obesity and prostate volume, four of which have
shown a significant positive association with BMI, waist circumference, and/or waist-
to-hip ratio.[18-21] These results suggest that BPH risk is positively associated with
obesity, in particular abdominal obesity, and that the metabolic abnormalities
associated with abdominal adiposity are likely to contribute to the pathogenesis of
BPH.
Although evidence from several studies indicates obesity is a risk factor for
BPH, relatively little is known about the mechanisms that underlie this association.
Obesity is characterized by enlarged depots of adipose tissue, which are metabolically
active and produce numerous hormones and proteins, collectively called adipokines.
Adipokines influence both local adipocyte biology as well as distant organ
systems[22,23]. Obesity is associated with higher concentrations of insulin[24],
leptin[25] and inflammatory cytokines[26], and lower concentrations of
adiponectin[27]; thus it is reasonable to hypothesize that obesity-related abnormalities
9
in the circulating concentrations of these factors may play a role in the pathogenesis of
BPH.
Adiponectin, Leptin, C-Peptide and BPH
Adiponectin
Adiponectin is a hormone produced exclusively by adipocytes which, unlike
most adipokines, is decreased in obesity[28]. Adiponectin plays a role in several
physiologic processes, including cell proliferation and apoptosis[29] and metabolism of
glucose and fatty acids[30], and there is evidence to support both direct and indirect
effects of adiponectin on BPH pathogenesis. Both benign and malignant human
prostate tissue express cell surface receptors for adiponectin (AdopoRl and R2)[31],
which are known to activate the AMPK, MAPK and NF-K signaling pathways[29]
involved in the regulation of cell growth, proliferation and apoptosis. In vitro,
adiponectin inhibits prostate cancer cell growth at physiologic concentrations and
suppresses dihydrotestosterone stimulated cell proliferation[32]. Thus, although the
effects of adiponectin on normal prostate cells are currently unknown, considerable
indirect evidence suggests that it may be anti-proliferative in prostate tissue.
In addition, adiponectin may indirectly affect risk of BPH through its effect on
insulin sensitivity. Adiponectin enhances insulin sensitivity by inhibiting both the
expression of hepatic gluconeogenic enzymes and endogenous production of
glucose[33]. In animal models, adiponectin administration lowers glucose
concentrations in normal mice and ameliorates hyperglycemia in ob/ob mice[34]. The
10
insulin-sensitizing effects of adiponectin are also seen in epidemiologic studies, which
find associations between low levels of adiponectin and several insulin-related
disorders including diabetes mellitus[35], insulin resistance[36], and metabolic
syndrome[37].
Only two small cross-sectional studies to date have investigated the association
between adiponectin and risk of BPH[38,39]; one (n=75) found a non-significant
decreased risk of prevalent BPH with increasing adiponectin concentrations [38] and
one (n=41) found no association[39].
Leptin
Leptin is an adipocyte-derived hormone that is secreted in proportion to adipose
tissue mass [25], and is elevated in obesity. Leptin plays an essential role in the
regulation of food intake and energy expenditure[22], and there is evidence to support
both direct and indirect effects of leptin on BPH pathogenesis. Leptin is thought to
play an important role in the development and maintenance of reproductive tissues,
including the prostate. [40] In addition, prostate tissue expresses high mRNA levels of
leptin receptor (Ob-R)[41], which has been shown to activate the phosphatidyl-inositol-
3-kinase (PI3-K) and cJUN NH2-terminal kinase (JNK) signaling pathways[42]
involved in the regulation of cell growth, proliferation and apoptosis. In vitro, leptin
inhibits the proliferation of androgen-independent prostate cancer cells[42]. Thus,
although the effects of leptin on normal prostate cells are currently unknown,
considerable indirect evidence suggests that it may be anti-proliferative in prostate
11
tissue.
In addition, leptin may indirectly affect risk of BPH through its effect on
autonomic nervous system activity and insulin sensitivity. Leptin increases
sympathetic nerve activity to the kidney and adrenal gland [43], which could result in
increased prostate growth. Increased sympathetic neural input to the prostate is
associated with prostate growth rate, the absence of which results in regression of the
prostate gland. [44] Furthermore, several epidemiologic studies report an association
between increased autonomic tone and symptomatic BPH [45-47], suggesting that
leptin-induced increases in sympathetic nervous system activity may affect BPH
development. High concentrations of leptin can also inhibit insulin secretion[48] and
increase insulin sensitivity[49]; however, obese individuals are often insulin resistant in
spite of high leptin levels. This suggests that effects of leptin depend largely on the
degree and sites of leptin resistance in obesity, and that obese people could be resistant
to the metabolic actions of leptin on insulin sensitivity, while not resistant to actions of
leptin on the sympathetic nervous system[50].
Several epidemiologic have examined the association between leptin
concentrations and BPH risk; however, only one study reported a non-significant
increased risk of LUTS (OR=1.20, 95%CI: -.95, 1.51) for the highest concentrations of
leptin[51]. All other studies have reported no association between leptin
concentrations and risk of BPH surgery. [52-54]
C-peptide
12
C-peptide is a peptide is a peptide that is made when pro-insulin is split into
insulin and C-peptide and thus provides a direct measure of insulin secretion from the
pancreas.[55] Although it is not an adipokines, it is strongly correlated with obesity
and insulin levels[56] and is often used as a marker of chronic hyperinsulinemia.
Insulin resistance is characterized by chronic elevations in circulating insulin
levels, which result in a reduced response to the effects of insulin on glucose uptake,
metabolism and storage.[57] There is good evidence suggesting that insulin resistance
promotes the development of BPH. In-vitro experiments have demonstrated that
insulin-like growth factor (IGF) is a critical factor in the development and proliferation
of the prostate. In addition, insulin receptors have been isolated in prostate
epithelium[58]. Evidence from epidemiologic studies also suggests that insulin
resistance plays a role in the pathogenesis of BPH. Aberrations in insulin control,
including a history of diabetes mellitus [20,59], and high levels of glycosylated
hemoglobin [59] and fasting glucose [20,59] are associated with an enlarged prostate
[20,60] and lower urinary tract symptoms (LUTS)[59].
No studies have examined the association between C-peptide and risk of BPH;
however, three studies have examined the association between fasting insulin levels
and risk of BPH. In two studies, fasting insulin concentrations were associated with
significantly larger prostate size[60,61], and in case-control studies, elevated fasting
insulin concentrations were associated with an increased risk of BPH surgery [54], and
symptomatic BPH[60].
13
Systemic Inflammation and BPH
There is consistent evidence from multiple disciplines suggesting that
inflammation plays a role in the etiology of BPH. Histologic studies have found acute
and/or chronic inflammation in up to 100% of BPH specimens [62-65]. In cross-
sectional studies, the presence of inflammatory infiltrates in prostate tissue is
associated with several measures of BPH including increased prostate volume[62,63],
more severe lower urinary tract symptoms (LUTS)[66], acute urinary retention
(AUR)[67,68], and epithelial cell proliferation[69]. Furthermore, in-situ studies have
found that expression of pro-inflammatory cytokines is increased in BPH tissue[70-73],
and in a rat prostate model, administration of immunostimulatory compounds induces
epithelial proliferation and hyperplastic lesions similar to BPH nodules[74]. In recent
prospective studies, participants with acute inflammation in biopsy specimens had a
greater risk of BPH progression (LUTS) and AUR[63]. Furthermore, in a large
population-based cohort of men, daily users of non-steroidal anti-inflammatory drugs
(NSAIDS) had a lower risk of several clinical measures of BPH (low maximum flow
rate, increased prostate volume and elevated PSA), and a lower risk of developing
moderate/severe LUTS[75].
The underlying causes of intra-prostatic inflammation remain unclear, though
several hypotheses have been proposed including: response to tissue damage caused by
infection[76] or other causes[77,78], or an autoimmune response[79]. It is also
plausible that systemic inflammation could contribute to the initiation or progression
14
within the prostate. Obesity and abdominal obesity, which are characterized by large
deposits of adipose tissue that produce an excess of inflammatory cytokines[26], are
well-established risk factors for BPH[12,13,54]. Thus, it is reasonable to hypothesize
that adipose-derived increases in circulating cytokine concentrations may influence
BPH risk.
Only two studies to date have investigated the association between circulating
levels of inflammatory markers and BPH risk. In a small, hospital-based case-control
study, men with histologically confirmed BPH had higher levels of interleukin-6 (IL-6)
than controls (1.9 vs.0.7 pg/ml).[80] More recently, Rohrmann et al. examined the
association between CRP and symptomatic BPH in the NHANES III cohort and found
that men with elevated circulating CRP levels (>0.30 mg/dl) were 1.47 times were
likely to have three or four LUTS than men with CRP concentration below the
detection limit. [81] Although the interpretation of these studies is limited by their
cross-sectional design, the consistency of the findings across several disciplines
supports our hypothesis that the association between obesity and BPH risk could be
mediated by systemic low-grade inflammation. Thus, further investigation using
prospective data would provide insight into the temporal sequence of this association.
BPH and Prostate Cancer
BPH and prostate cancer are very common urological conditions in older men
and historically, researchers believed the two conditions might be related. Originally
this hypothesis was based on evidence from autopsies where prostates with cancer had
15
a higher frequency of nodular hyperplasia than matched controls. [82-84] However,
differences in anatomical location and histology in the two conditions have cast doubt
on the possibility of an association. Nearly all BPH arises in the transition zone of the
prostate, while only 24% of prostate cancers occur in the transition zone, two-thirds in
the peripheral zone, and the rest in the central zone.[85] In addition, BPH is
characterized by hyperplasia of primarily stromal and, to a lesser extent, epithelial
cells, whereas prostate cancer involves hyperplasia of only the epithelium in the
glandular compartment of the prostate.[5,86] These differences fuel the current
opinion that BPH and prostate cancer are not linked in origin.[87]
Despite these differences, recent studies have identified a number of compelling
similarities between BPH and prostate cancer. Both conditions frequently coexist;
more than 80% of men with prostate cancer also have BPH and cancer is found
incidentally in a significant percentage (10-20%) of surgically removed BPH
specimens.[88,89] There are also several similarities between the two conditions: the
prevalence of both BPH and prostate cancer increase in parallel with age[88]; both
conditions require androgens for growth and development[88,90]; and both respond to
androgen-deprivation treatments. [91,92] Gene expression studies have also identified
similarities in genetic alterations between BPH and prostate cancer, particularly in the
growth regulatory genes[93]. In addition, an expanding body of evidence supports an
important role for inflammation in both[78,94].
Six previous studies [95-99] have examined the association between BPH and
16
prostate cancer risk, three of which reported an increased risk of prostate cancer
associated with BPH. [95,96] However, bias is a likely explanation of all the
previously published studies showing an association between BPH and prostate cancer.
The largest single source of bias in these studies is due to the increased likelihood of
prostate cancer detection in men with symptomatic urinary disease. Symptoms of BPH
can increase the detection rate of prostate cancer by increasing the number and extent
of urologic examinations, [100] and treatment of BPH via trans-uretheral resection of
the prostate (TURP) can result in the detection of incidental, latent prostate
cancer.[101] In addition, BPH can increase prostate specific antigen (PSA) levels,
thereby increasing the likelihood of detecting unsuspected prostate cancer. [102] BPH
also causes an enlargement of the prostate gland which can make it more difficult to
detect cancer. [103]
The Prostate Cancer Prevention Trial (PCPT) offers a unique opportunity to
investigate the association between BPH and prostate cancer. In the PCPT, all men
were subject to the same diagnostic scrutiny for prostate disease, regardless of
symptomatology; every participant received an annual DRE and PSA test, and
completed an assessment of lower urinary tract symptoms. Furthermore, all men who
were not diagnosed with interim prostate cancer were asked to undergo an end-of-study
biopsy to confirm the absence or presence of prostate cancer.
Specific Aims of the Dissertation
The overall goals of this dissertation are to understand the pathways through
17
which obesity increases the risk of symptomatic BPH, and to determine whether
symptomatic BPH is associated with risk of subsequent prostate cancer. Specifically, I
investigate whether obesity-related abnormalities in circulating concentrations of
adiponectin, leptin and C-peptide are associated with obesity, and whether systemic
concentrations of inflammatory markers elevated in obesity are involved in the
development of symptomatic BPH. I also examine the association between
symptomatic BPH and the risk of subsequent prostate cancer.
Data are from the placebo group participants in the Prostate Cancer Prevention
Trial (PCPT), a 7-year randomized, double-blinded trial in 18,884 men aged 55-70 to
test the drug finasteride (Proscar®) for the primary prevention of prostate cancer.
There assessment of prostate-related disease in PCPT was rigorous and complete,
including annual cancer screening (serum prostate specific antigen and digital rectal
exam) and assessment of lower urinary tract symptoms, quarterly interviews on
prostate disease-related diagnoses, surgical treatments and prescription drugs use, and a
prostate biopsy completed at the end of the study.
The specific aims of this dissertation are the following:
1. To test whether plasma concentrations of two adipokines, adiponectin and
leptin, and C-peptide, a measure of insulin secretion, affect the risk of
symptomatic BPH.
2. To test whether systemic concentrations of several inflammatory markers
elevated in obesity, including tumor necrosis factor-alpha (TNF-a), TNF
18
receptors I and II (sTNF-RI and II), interleukin-6 (IL-6), interferon-gamma
(IFN-y), and C -reactive protein (CRP) affect the risk of symptomatic BPH.
3. To test whether symptomatic BPH is associated with an increased risk of
prostate cancer.
19
Chapter 1
Serum Adiponectin, C-Peptide and Leptin and Risk of
Symptomatic Benign Prostatic Hyperplasia
Abstract
Recent epidemiologic studies have identified obesity as a risk factor for benign
prostatic hyperplasia (BPH). We examined whether adiponectin, leptin and C-peptide
were associated with incident, symptomatic BPH and whether these factors mediate the
relationship between obesity and BPH risk. Data are from Prostate Cancer Prevention
Trial placebo arm participants who were free of BPH at baseline. Incident BPH
(n=698) was defined as treatment, two International Prostate Symptom Score (IPSS)
values >14, or an increase of >5 in IPSS from baseline documented on at least two
occasions plus at least one score >12. Controls (n=709) were selected from men
reporting no BPH treatment or IPSS >7 during the seven-year trial. Baseline serum
was analyzed for adiponectin, C-peptide, and leptin concentrations. Neither C-peptide
nor leptin was associated with BPH risk. The odds ratio [95% CI] contrasting highest
to lowest quartiles of adiponectin was 0.65[0.47, 0.87] P
t
rend
=
0.004. Findings differed
between levels of physical activity: there was a strong inverse association between
adiponectin and BPH among moderately/very active men OR=0.43[0.29, 0.63], and no
association among sedentary/minimally active men OR=0.92[0.65, 1.30]
Pinteraction=0.005. Adiponectin concentrations explained only a moderate amount of the
relationship between obesity and BPH risk. High adiponectin concentrations were
associated with reduced risk of incident, symptomatic BPH. This association was
20
limited to moderately/very active men; suggesting the relationship between obesity and
BPH involves a complex interaction between factors affecting glucose uptake and
insulin sensitivity. However, adiponectin is likely not the only mechanism through
which obesity affects BPH risk.
Introduction
Benign prostatic hyperplasia (BPH) is one of the most common medical
conditions in middle-aged and older males, affecting 40 to 50 percent of men by age 50
and up to 80 percent of men by age 70[3,104]. Despite the high prevalence of BPH, its
pathogenesis remains unclear, and few modifiable risk factors have been established.
Recent epidemiologic studies have identified obesity as a risk factor for BPH
[12,13,54]. The largest of these studies, from the Prostate Cancer Prevention Trial
(PCPT)[12] and the Health Professionals Follow-up Study cohorts[13], both found that
obesity, in particular abdominal obesity, was associated with a 16-43% significant
increase in risk of symptomatic BPH.
Relatively little is known about the mechanisms that underlie the association of
obesity and BPH risk. Obesity is characterized by enlarged depots of adipose tissue,
which are metabolically active and secrete numerous hormones and proteins that
influence local adipocyte biology as well as distant organ systems[22]. Obesity is
associated with higher serum insulin[24] and leptin[25], and lower adiponectin[28];
thus, it is reasonable to hypothesize that obesity-related abnormalities in the circulating
concentrations of these factors may play a role in the pathogenesis of BPH.
21
Here we give results of a nested case-control study examining risk of incident,
symptomatic BPH among placebo-arm participants in the PCPT. This report examines
whether concentrations of two adipokines, adiponectin and leptin, and C-peptide, a
measure of insulin secretion, affect the risk of symptomatic BPH. We also evaluated
whether these factors mediate the relationship between obesity and risk of BPH.
Methods
Data are from the PCPT, a randomized, placebo-controlled trial testing whether
finasteride, a 5a-reductase inhibitor, could reduce the 7-year period prevalence of
prostate cancer. Details regarding study design and participant characteristics have
been described previously[92]. Briefly, 18,880 men age 55 years and older with a
normal digital rectal exam, prostate-specific antigen level of 3 ng/ml or below, no
history of prostate cancer or other clinically significant coexisting conditions, and no
severe BPH symptoms (International Prostate Symptom Score (IPSS) of 20 or higher),
were randomized to receive finasteride (5 mg/day) or placebo.
Participants for this nested case-control study were drawn from the 9,457 men
randomized to the placebo arm of the PCPT. Analyses are restricted to the placebo arm
because finasteride both prevents and treats BPH[92]. Exclusion criteria included men
who, at baseline, had medical or surgical treatment for BPH (n=701), a self-reported
history of BPH (n=l,904), or an IPSS >7 (n=l,820). In addition, men using steroid
hormones at any time (n=61) were excluded, leaving 4,971 men eligible for this study.
Extensive medical data, including self-report of physician diagnosis of and
22
treatment for BPH were collected at the baseline visit, each clinic visit (6- and 12-
month), and at every phone contact (3- and 9-month) between clinic visits. At
recruitment (3 months prior to randomization), randomization and each 12-month
clinic visit, participants completed the 7-item IPSS as a self-administered
questionnaire. Clinic staff measured height, weight, waist and hip circumference at
randomization, and body mass index (BMI) was calculated as weight (kg) divided by
height
2
(m). Age, race/ethnicity, physical activity (type, frequency, duration, pace, and
intensity), alcohol consumption (type, frequency and dose) and history of smoking
were collected at baseline using self-administered questionnaires. Physical activity
was categorized into 4 levels; sedentary (low activity index (calculated as the multiple
of frequency, duration and pace indices) and participated in intense activities less than
once a week), light (low activity index and participated in intense activities four times a
week or less), moderate (moderate or high activity index and participated in intense
activities four times a week), and very active (high activity index and participated in
intense activities five or more times a week).
Incident BPH was defined either as a report of treatment for or development of
significant lower urinary tract symptoms. Treatments included use of a-blockers,
finasteride or any surgical intervention (transurethral prostatectomy, balloon dilation or
laser prostatectomy). Development of significant symptoms was defined as either (1)
two IPSS scores >14; or (2) an increase of >5 in IPSS from baseline documented on at
least two occasions plus at least one score >12. There were a total of 727 incident BPH
23
cases, which were further categorized into clinical BPH (322 cases defined by medical
or surgical treatment and 105 cases defined by two IPSS scores of 15 or higher) and
symptomatic progressive BPH (300 cases defined by two IPSS scores at least 5 units
higher than baseline plus at least one score >12). Men who reported transient
elevations in IPSS or a physician diagnosis of BPH in the absence of symptoms or
treatment were not included as cases.
Controls were drawn from the 1,497 men who, during seven years of
observation, had no more than two missing IPSS, no single IPSS >7, no surgical or
medical treatment for BPH, and no report of a diagnosis of BPH. From this sample, we
selected all men aged 70 years and over and all non-Caucasian men, in order to
maximize statistical power when examining these subgroups. Remaining controls were
randomly selected to yield a total control sample (n=727) that was frequency matched
to the age distribution (in 5 year age groups) of cases. To minimize the influence of
diabetes on either the exposure (serum adiponectin, C-peptide or leptin concentrations)
or outcome (BPH risk), we further excluded men with a self-reported history of
diabetes (diagnosis or treatment) at baseline (n=6 cases and n=2 controls).
Non-fasting blood samples were drawn at recruitment into a 7 ml EDTA
vacutainer tube and shipped overnight to a central storage facility, where they were
centrifuged, aliquoted and stored at -70 °C until analysis. Serum concentrations of
adiponectin, C-peptide, and leptin were quantified using a multiplex sandwich enzyme-
linked immunosorbent assay at Pierce Biotechnology Inc., Woburn, MA. Samples
24
were run in a series of dilutions, and plates were imaged using the SearchLight® Black
Ice Imaging system, and images were analyzed using SearchLight® Array Analyst
software. Serum was not available from 39 men, and analyses were completed using
698 cases and 709 controls.
Descriptive statistics were used to characterize the study sample and
distribution of adiponectin, C-peptide, and leptin concentrations in cases and controls.
Serum concentrations of analytes were categorized into quartiles based on the
distribution in the controls, and unconditional logistic regression was used to calculate
relative odds ratios (OR, referred to as risk) and their 95% confidence intervals for risk
of BPH. All models were adjusted for age at baseline (continuous), BMI (continuous),
race (white, other), and alcohol consumption (continuous). Controlling for other
factors associated with BPH risk in this sample, including baseline IPSS[12], waist to
hip ratio[12], insulin-like growth factors (IGF-1, and IGFBP-3)[105], steroid hormones
(testosterone, 3-a-diol gulcaronide, testosterone:3-a-diol gulcaronide ratio, and serum
hormone binding globulin)[106] and dietary intake (fat, protein, red meat, and
vegetable servings)[107], did not affect risk estimates and were not included in the
final models. Tests for linear trend across quartiles were performed by using an ordinal
variable corresponding to rank from lowest to highest category[108].
Additional analyses were completed stratified by age, physical activity, alcohol
intake, and BMI. To increase the stability across stratified estimates, serum
concentrations of adipokines and C-peptide were categorized into tertiles based on the
25
distribution among controls. Formal interaction tests were based on p-values for the
interaction term of the adipokine or C-peptide concentration rank (from lowest to
highest) times a dummy variable for physical activity (sedentary/light versus
moderately/very active), alcohol (<1 versus 1+ drink/day) or BMI (<25 versus >25).
We also considered whether the previously-reported finding of an increased BPH risk
in obese men[12] was mediated by adiponectin, C-peptide or leptin by examining the
association of obesity with BPH risk both unadjusted and adjusted for these analytes.
All p-values were two-sided and considered statistically significant at p<0.05.
Statistical analyses were conducted using SAS (version 9.1 Cary, NC).
Results
Distributions of demographic, anthropometric, and health-related variables are
given in Table 1. Participants were mostly white, overweight or obese, and
nonsmokers. Compared with controls, men with BPH consumed less alcohol and had a
higher IPSS at baseline. Table 2 gives mean/median values and 5 and 95 percentile
distributions of adipokines and C-peptide in cases and controls. Compared with
controls, cases had a lower concentration of adiponectin at baseline (Table 2).
Table 3 gives adjusted odds ratios for risk of BPH associated with serum
adiponectin, C-peptide, and leptin concentrations. In models adjusted for matching
variables (age, race), BMI and alcohol, men in the highest quartile of serum
adiponectin had a 35% lower risk of BPH with a monotonic and statistically significant
linear trend (Ptrend=0.004). Neither C-peptide nor leptin were associated with risk of
26
BPH. There were no substantial differences in results when BPH outcomes (clinical
versus symptomatic progression, Table 1.4) were considered separately, when analyses
excluded men diagnosed with prostate cancer (n=104 cases and n=l 15 controls, Table
1.5), or by lag time between blood collection and BPH event (1-3 years versus 4-7
years, Table 1.6).
Several analyses were completed to explore possible effect modification.
Associations between BPH and serum adiponectin, C-peptide, and leptin were similar
across subgroups stratified by BMI, and alcohol intake (Table 7). However, for strata
characterized by level of physical activity, there was no association of adiponectin with
BPH risk among sedentary/minimally active men, and a strong inverse association
among men who were moderately/very active (P interaction^.005, Table 7).
In previous analyses in this cohort, obesity was associated with a increased risk
of BPH [12]; therefore, we completed additional analyses to examine whether
adipokines or C-peptide mediate the association between obesity and risk of BPH.
Compared to men who are normal weight, overweight and obese men have a 24%
(OR=1.24, 95%CI 0.96-1.60) and 36% (OR=1.36, 95%CI 1.00-1.85) increased risk of
BPH, respectively (Ptrend=0.04). Control for adiponectin reduced the overall
association between obesity and BPH risk by 25% to 1.18 (95%CI, 0.91 -1.52) for
overweight, and by 28% to 1.26 (95%CI, 0.92-1.72) for obese (Ptrend=0.14). Control
for C-peptide and leptin had no affect on the overall association between obesity and
BPH risk (Table 1.8).
27
Discussion
In this nested case-control study there was an inverse dose response association
between serum adiponectin concentration and risk of symptomatic BPH. This
association did not differ by BMI, alcohol intake, type of BPH diagnosis or time
between blood draw and BPH diagnosis; however, it was modified by physical activity
level. There was no association between adiponectin and BPH among men who were
sedentary/minimally active, and a strong inverse association among men who were
moderately/very active. Neither serum C-peptide nor leptin were associated with BPH
risk.
Few epidemiologic studies have investigated the association between
adiponectin, C-peptide or leptin and risk of BPH [38,39,52,54,60]; most were limited
by small numbers of BPH cases and use of a cross-sectional study design. Two cross-
sectional studies examined the association between leptin concentrations and risk of
BPH surgery[54] or clinical BPH[52], and consistent with the findings of this
investigation, neither found an association between leptin concentrations and risk of
prevalent BPH. No studies have examined the association between C-peptide and risk
of BPH; however several have examined the relationship between fasting insulin levels
and risk of BPH. The results of these studies, which are inconsistent with our findings,
suggest an association between insulin levels and BPH risk. In cross-sectional studies,
fasting insulin concentrations were associated with a significantly larger prostate
size[20,60], and in case-control studies, elevated fasting insulin concentrations were
28
associated with an increased risk of BPH surgery[54] and symptomatic BPH[60]. Only
two small cross-sectional studies have investigated the association between adiponectin
and risk of BPH[38,39]; one (n=75) found a non-significant decreased risk of prevalent
BPH with increasing adiponectin concentrations [38] and one (n=41) found no
association[39].
Adiponectin is a hormone produced exclusively by adipocytes which, unlike
most adipokines, is decreased in obesity[28]. Adiponectin plays a role in several
physiologic processes, including cell proliferation and apoptosis[29] and metabolism of
glucose and fatty acids[30], and there is evidence to support both direct and indirect
effects of adiponectin on BPH pathogenesis. Both benign and malignant human
prostate tissue express cell surface receptors for adiponectin (AdopoRl and R2)[31],
which are known to activate the AMPK, MAPK and NF-K signaling pathways[29]
involved in the regulation of cell growth, proliferation and apoptosis. In vitro,
adiponectin inhibits prostate cancer cell growth at physiologic concentrations and
suppresses dihydrotestosterone stimulated cell proliferation[32]. Thus, although the
effects of adiponectin on normal prostate cells are currently unknown, considerable
indirect evidence suggests that it may be anti-proliferative in prostate tissue.
In addition, adiponectin may indirectly affect risk of BPH through its effect on
insulin sensitivity. Adiponectin enhances insulin sensitivity by inhibiting both the
expression of hepatic gluconeogenic enzymes and endogenous production of
glucose[33]. In animal models, adiponectin administration lowers glucose
29
concentrations in normal mice and ameliorates hyperglycemia in ob/ob mice[34]. The
insulin-sensitizing effects of adiponectin are also seen in epidemiologic studies, which
find associations between low levels of adiponectin and several insulin-related
disorders including diabetes mellitus[35], insulin resistance[36], and metabolic
syndrome[37]. There is good evidence that insulin resistance promotes the
development of BPH. Aberrations in insulin control, including a history of diabetes
mellitus[20,59], and high levels of glycosylated hemoglobin[59] and fasting glucose
[20,59] are associated with an enlarged prostate[20,60] and lower urinary tract
symptoms(LUTS)[59]. Thus, the evidence is compelling that adiponectin, through
increased insulin sensitivity, could reduce BPH risk.
Given our finding of a strong effect for adiponectin on BPH risk and the
established effects of adiponectin on insulin sensitivity, the lack of association between
C-peptide and BPH in this study was somewhat unexpected. It is possible that an
association between C-peptide and BPH does exist, but was not detectable in this study
due to increased measurement variance resulting from the use of non-fasting serum.
Alternatively, C-peptide is a direct marker of endogenous insulin secretion[55],
whereas adiponectin, whose insulin sensitizing effects are independent of insulin
levels[34], is a marker of insulin sensitivity[36]. Therefore, insulin sensitivity rather
than insulin secretion may be the important factor in development of BPH.
Our finding that the inverse association between adiponectin and BPH is
limited to men with higher levels of physical activity is difficult to interpret. One
30
possible interpretation is that physical activity and adiponectin affect BPH risk through
complementary mechanisms. Both physical activity and adiponectin improve insulin
sensitivity; adiponectin enhances insulin action by suppressing hepatic glucose
production[34] and increasing glucose uptake and fatty acid oxidation in muscle[30],
whereas physical activity improves insulin sensitivity primarily by inducing the uptake
of glucose in skeletal muscle[109]. In addition, physical activity may enhance the
activity of adiponectin by increasing circulating adiponectin levels and adiponectin
receptor expression in muscle tissue[l 10]. Perhaps both mechanisms are required in
order to see a beneficial effect.
In this PCPT subsample we observed an association between overall
obesity(BMI) and BPH risk, as was found in the larger PCPT study of obesity and
BPH[12]. However, when examining whether the association between obesity and
BPH risk was mediated by adiponectin, adjustment for adiponectin resulted in only a
modest attenuation of the overall risk estimate, indicating that adiponectin is not the
only mechanism through which obesity affects BPH risk.
This study has several strengths. It is a prospective analysis in which men with
a history of BPH or BPH symptoms were excluded at the time of the baseline blood
draw. We had extensive data on covariates and could control for many factors
associated with BPH risk. Of special note is baseline IPSS, which was strongly
associated with incidental BPH[12]; however, control for baseline IPSS had no
appreciable effect on the association of adiponectin and BPH risk (OR for Q4 vs.
31
Ql=0.67, 95%CI 0.48, 0.93). This study uses a rigorous definition of BPH that
captures all current medical treatments as well as LUTS, which were collected annually
using a standardized and well validated self-administered questionnaire. This is in
contrast to most previous studies which relied on prostate size, which correlates poorly
with symptoms[l 11], or BPH surgery alone as an endpoint, which may be biased
because this definition cannot separate factors associated with surgical treatment (e.g.
health insurance, age, co-morbidities) from those that predict severe BPH.
There are several limitations to this study. First, our definition of BPH by
LUTS can not distinguish between LUTS due to BPH and LUTS due to other urologic,
neurologic or bladder conditions[l 12]. We used the age-specific prevalences of 'other
conditions' associated with LUTS within the Olmstead County cohort (13.5%)[113] to
estimate a prevalence of these conditions (6.5%) in the PCPT study population (lower
because the PCPT excludes men with a history of prostate surgery). This level of non-
specificity is unlikely to explain our study results. Second, in order to maximize the
contrast between groups we deliberately selected a control group with a definitive
absence of symptoms associated with BPH. Thus, men who developed mild to
moderate symptoms were excluded from this study. It is unclear whether the
development of mild/moderate symptoms is indicative of an intermediate form of
symptomatic BPH, though the inclusion of these men would likely have attenuated our
results. Third, a single measurement of leptin may be susceptible to short-term
variations; thus the results from our study may have been biased resulting in an under-
32
estimation of the true relationship between serum leptin concentrations and BPH risk.
Fourth, bloods were non-fasting and drawn at all times during the day, which probably
increased variability in adipokine and C-peptide concentrations. However, we have no
reason to believe that the distribution of time since last meal or time of blood draw
differed markedly between cases and controls; therefore, this error is unlikely to
introduce differential bias and would only attenuate the associations with BPH.
In conclusion, we found that higher adiponectin concentrations were associated
with a reduced risk of symptomatic BPH; however, only among men who were
moderately/very active. This suggests that the relationship between obesity and BPH
risk involves a complex interaction between factors that affect glucose uptake and
insulin sensitivity. Adiponectin explained only a moderate proportion of the
association between obesity and BPH risk, and thus it is not the only mechanism
through which obesity affects risk of BPH. It is possible that several other obesity-
related metabolic abnormalities, including alterations in steroid hormone levels, and
chronic, low-level inflammation could also contribute to the development of BPH.
Future studies will be needed to confirm our finding and assess the role of other
potential mediating pathways.
33
Table 1.1. Demographic and lifestyle characteristics of the study population
Age
55-59
60-64
65-69
70+
Race/Ethnicity
White
Other
Waist to hip ratio
<0.95
0.95-0.99
1.00-1.04
1.05+
Body Mass Index (kg/m
2
)
Normal (<25)
Overweight (25-29.9)
Obese (>30)
Alcohol Consumption (grams/day^
<0.25
0.25 -1.74
1.75-12.9
13-26. 9
>27
Physical activity
Sedentary
Light activity
Moderate activity
Very active
Smoking status
Yes (current)
No (former/never)
Baseline IPSS
1-3
4-5
6-7
Cases
n
209
197
187
105
647
51
261
220
125
18
164
362
161
1
198
92
241
100
66
112
312
199
72
52
645
309
281
108
Benign Prostatic Hyperplasia (BPH)
(N=698)
b
(%)
(30.0)
(28.2)
(26.8)
(15.0)
(92.7)
(7.3)
(37.4)
(31.5)
(17.9)
(2.6)
(23.5)
(51.9)
(23.1)
(28.4)
(13.2)
(34.5)
(14.3)
(9.5)
(16.1)
(44.7)
(28.5)
(10.3)
(7.5)
(92.4)
(44.3)
(40.3)
(15.5)
Controls (N=709)
b
n
213
198
186
112
657
52
306
229
113
21
200
360
146
155
128
238
109
79
102
297
234
74
41
668
552
140
17
(%)
(30.1)
(27.9)
(26.2)
(15.8)
(92.7)
(7.3)
(43.2)
(32.3)
(15.9)
(3.0)
(28.2)
(50.8)
(20.6)
(21.9)
(18.0)
(33.6)
(15.4)
(11.1)
(14.4)
(41.9)
(33.0)
(10.4)
(5.8)
(94.2)
(77.9)
(19.8)
(2.4)
P
c
0.98
0.88
0.28
0.14
0.02
0.30
0.22
<0.01
"Excludes 23 cases and 16 controls with missing adipokine and c-peptide data, and 6 cases and 2
controls with a history of diabetes at baseline.
''Numbers may not add up to the total number of cases/controls and percentages may not add up to 100
due to missing data.
c
P from chi-square test
34
Table 1.2. Serum concentration of adipokines and C-peptide for cases and
controls
Benign Prostatic Hyperplasia (BPH)
Cases (N=698) Controls (N=709) P
Adiponectin (ug/ml)
Geometric mean (95%CI) 4.86(3.63,6.49) 5.42(4.06,7.24) <0.01
Median 4.84 5.41 <0.01
a
5
t h
- 95
th
percentiles 0.64-25.60 0.80-26.14
C-peptide (ng/mi)
Geometric mean (95%CI) 1.30(0.95,1.79) 1.31(0.92,1.70) 0.82
Median 1.35 1.33 0.96
a
5
t h
- 95
th
percentiles 0.18-7.28 0.19-5.67
Leptin (ng/ml)
Geometric mean (95%CI) 4.72(3.42,6.55) 4.56(3.25,6.55) 0.41
Median 4.59 4.74 0.4 l
a
5
t h
- 95
th
percentiles 0.76-34.27 0.62-49.46
P from Wilcoxon rank-sum test
35
Table 1.3. Association of adipokines and C-peptide on BPH risk
Adiponectin, (ug/ml)
Qi
Q2
Q3
Q4
P for trend
C-peptide, (ng/ml)
Qi
Q2
Q3
Q4
P for trend
Leptin, (ng/ml)
Qi
Q2
Q3
Q4
P for trend
No. cases/
controls
222/176
177/178
161/177
138/178
191/177
151/177
185/177
171/178
141/177
223/177
146/177
188/178
Adjusted
1
OR
1.00
0.78
0.71
0.60
(95% CI)
(0.59,1.04)
(0.53,0.96)
(0.45,0.82)
0.0008
1.00
0.79
0.96
0.89
(
1.00
1.59
1.03
1.33
(
(0.58,1.06)
(0.72,1.29)
(0.66,1.19)
3.72
(1.18,2.14)
(0.76,1.41)
(0.98,1.79)
3.47
Fully Adjusted
2
OR
1.00
0.80
0.75
0.65
1.00
0.74
0.90
0.80
1.00
1.48
0.91
1.05
(95% CI)
(0.60,1.07)
(0.56,1.01)
(0.47,0.87)
0.004
(0.55,1.00)
(0.67,1.22)
(0.59,1.08)
0.31
(1.09,2.01)
(0.66,1.27)
(0.73,1.50)
0.48
Quartile outpoints: C-peptide: 0.87, 1.33, 2.09; Leptin: 2.68, 4.74, 7.39; Adiponectin: 3.78, 5.41, 7.71
Adjusted for matching covariates only (age at baseline and race)
2
Adjusted for matching covariates, and body mass index(linear), and alcohol intake (grams/day)
36
Table 1.4. Associations of Adiponectin, C-peptide and Leptin (quartiles) with risk
of BPH categorized as clinical or symptomatic progression BPH.
Symptomatic Progression
BPH
Clinical BPH
No. cases/
controls
OR (95% CI)'
No. cases/
controls
OR (95% CI)
1
Adiponectin, (ug/ml)
Qi
Q2
Q3
Q4
P for trend
C-peptide, (ng/ml)
Qi
Q2
Q3
Q4
P for trend
Leptin, (ng/ml)
Qi
Q2
Q3
Q4
P for trend
129/176
100/177
94/175
80/178
103/177
89/177
113/175
98/177
74/176
131/176
90/176
108/178
1.00
0.80(0.57-1.12)
0.76(0.54-1.07)
0.66 (0.46-0.94)
0.02
P
2
-
•* difference
1.00
0.83(0.58-1.18)
1.02(0.73-1.45)
0.86(0.60-1.23)
0.66
P
2
-
1
difference
1.00
1.69(1.18-2.42)
1.07(0.72-1.59)
1.16(0.76-1.78)
0.85
P
2
-
1
difference
90/176
mm
' 64/175
57/178
= 0.78
87/177
57/177
70/175
69/177
= 0.38
66/176
87/176
54/176
76/178
= 0.43
1.00
0.80(0.55-1.16)
0.71 (0.48-1.05)
0.64 (0.43-0.95)
0.02
1.00
0.63 (0.42-0.94)
0.76(0.52-1.12)
0.72(0.49-1.07)
0.18
1.00
1.25(0.85-1.84)
0.73(0.47-1.14)
0.92(0.58-1.46)
0.29
Adjusted for age at baseline, race, BMI, alcohol,
2
' p-difference is calculated as the difference in risk between clinical and symptomatic
progression BPH
37
Table 1.5. Association of adipokines and C-peptide on BPH
risk, excluding men with prostate cancer
Adiponectin, (ug/ml)
Ql (ref)
Q2
Q3
Q4
P for trend
C-peptide, (ng/ml)
Ql (ref)
Q2
Q3
Q4
P for trend
Leptin, (ng/ml)
Ql (ref)
Q2
Q3
Q4
P for trend
No. cases/
controls
191/156
158/144
144/153
110/147
160/150
135/148
152/147
156/155
115/145
201/146
123/153
164/156
Fully Adjusted
1
OR (95% CI)
1.00
0.91
0.81
0.64
(0.66-1.25)
(0.59-1.12)
(0.46-0.89)
0.0008
1.00
0.79
0.90
0.84
1.00
1.64
0.91
1.07
(0.57-1.10)
(0.65-1.25)
(0.61-1.17)
0.45
(1.18-2.29)
(0.63-1.30)
(0.73-1.59)
0.47
Adjusted for matching covariates, and body mass index(linear),
and alcohol intake (grams/day)
38
Table 1.6. Associations of Adiponectin, C-peptide and Leptin (quartiles) with risk
of BPH categorized as BPH cases captured in years 1-3 or 4-7.
Cases captured
years 1-3
Cases captured
years 4-7
No. cases/
controls
OR (95% CI)'
No. cases/
controls
OR (95% CI)
1
Adiponectin, (ug/ml)
Qi
Q2
Q3
Q4
P for trend
C-peptide, (ng/ml)
Qi
Q2
Q3
Q4
P for trend
Leptin, (ng/ml)
Qi
Q2
Q3
Q4
P for trend
35/176
35/177
28/175
21/178
28/177
23/177
36/175
32/177
17/176
38/176
33/176
31/178
1.00
1.03(0.61-1.72)
0.83(0.48-1.43)
0.64(0.35-1.15)
0.11
p
2
r
difference
1.00
0.80(0.44-1.44)
1.22(0.71-2.10)
1.05(0.60-1.85)
0.55
p
2
-
r
difference
1.00
2.15(1.16-3.98)
1.76 (0.92-3.36)
1.51 (0.74-3.10)
0.49
p
2
-
r
difference
184/176
137/177
130/175
116/178
= 0.93
162/177
123/177
147/175
135/177
= 0.17
123/176
180/176
111/176
153/178
= 0.20
1.00
0.76(0.56-1.03)
0.72 (0.53-0.99)
0.65 (0.47-0.90)
0.008
1.00
0.80(0.44-1.44)
1.22(0.71-2.10)
1.05(0.60-1.85)
0.55
1.00
1.39(1.01-1.90)
0.80(0.56-1.13)
0.98(0.68-1.43)
0.29
Adjusted for age at baseline, race, BMI, alcohol,
2
p-difference is calculated as the difference in risk between clinical and symptomatic
progression BPH
39
Table 1.7. Association of adipokines and C-peptide with BPH risk, stratified by physical
activity, alcohol consumption and body mass index."
Adiponectin, (ug/ml)
Tl
T2
T3
C-peptide, (ng/ml)
Tl
T2
T3
Leptin, (ng/ml)
Tl
T2
T3
Adiponectin, (ug/ml)
Tl
T2
T3
C-peptide, (ng/ml)
Tl
T2
T3
Leptin, (ng/ml)
Tl
T2
T3
Sedentary/Minimally Active
No. cases/
controls
173/141
120/142
124/114
134/119
125/127
158/151
115/122
149/129
153/146
<1
No. cases/
controls
170/153
154/177
197/188
223/184
164/169
134/165
162/168
193/167
166/183
OR (95% CI)
b
1.00
0.71(0.51,0.99)
0.92 (0.65, 1.30)
- *i nt eract i on^^ • v\JZ>
1.00
0.84(0.59,1.20)
0.85 (0.60, 1.20)
P. . =f) 94
1
interaction V. Z - T
1.00
1.15(0.81,1.63)
0.93(0.63,1.35)
p. . =n S7
-* interaction
u
- - ' '
drink/day
c
OR (95% CI)
2
1.00
0.81 (0.61, 1.09)
0.69(0.51,0.93)
p. . =n s"*
-* interaction ^J-~>~>
1.00
0.75(0.55, 1.03)
0.88(0.65,1.19)
P. . = 0 ^Q
-* interaction v.-j?
1.00
1.10(0.81, 1.50)
0.77(0.54, 1.08)
p. . =n 07
x
interaction
u
*
u
'
Moderately/Very Active
No. cases/
controls
117/95
90/93
60/119
101/116
80/107
86/84
89/111
111/105
67/91
1+
No. cases/
controls
67/83
51/58
47/47
68/52
47/66
50/70
42/67
69/67
54/54
OR (95% CI)
b
1.03 (0.72, 1.46)
0.81 (0.56, 1.16)
0.43 (0.29, 0.63)
0.79(0.55,1.13)
0.63 (0.43, 0.92)
0.85(0.58, 1.27)
0.86(0.59,1.26)
1.05(0.72,1.53)
0.65(0.42,1.00)
drink/day
c
OR (95%o CI)
2
1.16(0.71,1.89)
0.64(0.38,1.10)
0.65 (0.39, 1.10)
0.83 (0.51, 1.35)
0.86(0.51,1.44)
0.97(0.54, 1.72)
0.73 (0.43, 1.24)
1.13(0.68,1.88)
0.98 (0.57, 1.70)
40
Table 1.7. Continued. Association of adipokines and C-peptide with BPH risk, stratified
by physical activity, alcohol consumption and body mass index."
Adiponectin, (ug/ml)
Tl
T2
T3
C-peptide, (ng/ml)
Tl
T2
T3
Leptin, (ng/ml)
Tl
T2
T3
No. cases/
controls
44/51
59/68
61/81
78/81
49/70
37/49
98/125
54/53
15/22
BMI < 25
OR (95% CI)
2
1.00
1.00(0.58,1.70)
0.90(0.53,1.52)
P. . =0 XI
1
interaction
yJ
--
J
'
1.00
0.71 (0.43, 1.14)
0.75 (0.44, 1.28)
P. . =() 7T, 1
interaction
w
-^—'
1.00
1.32(0.83,2.10)
0.87 (0.43, 1.78)
p . . =0 4?
J
interaction v.-it.
No. cases/
controls
247/185
152/167
123/154
159/155
156/165
207/186
109/110
208/181
205/215
BMI > 25
OR (95% CI)
2
1.42(0.87,2.32)
0.98(0.59,1.61)
0.86(0.52, 1.43)
0.97(0.64,1.48)
0.86(0.56,1.33)
1.00(0.64,1.55)
1.16(0.77,1.74)
1.30(0.88,1.91)
0.98 (0.63, 1.53)
"Adjusted for age, race, body mass index(linear), and alcohol intake (grams/day)
b
Joint effects models, in which the odds ratios for interactions are reported using a single reference group
c
1 drink per day is equivalent to 13 grams of alcohol per day
41
Table 1.8. Main effects of BMI on BPH risk, adjusted for adipokines and C-peptide
Model 1 Model 2 Model 3 Model 4
Normal weight 1.00 1.00 1.00 1.00
Overweight 1.24(0.96-1.60) 1.18(0.91-1.52) 1.27(0.97,1.66) 1.26(0.97,1.62)
Obese 1.36(1.00,1.85) 1.26(0.92-1.72) 1.43(0.99,2.07) 1.40(1.02,1.92
P for trend O04 (U4 O05 0.03
Model 1: Adjusted for age, race, alcohol
Model 2: Adjusted for age, race, alcohol and adiponectin
Model 3: Adjusted for age, race, alcohol and leptin
Model 4: Adjusted for age, race, alcohol and C-peptide
42
Chapter 2
Biomarkers of Systemic Inflammation and Risk of Incident,
Symptomatic Benign Prostatic Hyperplasia
Abstract
The authors conducted a nested case-control study of serum inflammatory
markers and risk of symptomatic benign prostatic hyperplasia (BPH), using data from
the placebo-arm of the Prostate Cancer Prevention Trial. Incident BPH (n=676) was
defined as treatment, report of 2 International Prostate Symptom Score (IPSS)
values>14, or 2 increases of >5 from baseline IPSS with at least one value>12.
Controls (n=683) were men who reported no BPH treatment or IPSS>7 over the 7-year
trial. Baseline serum was analyzed for C-reactive protein (CRP), tumor necrosis
factor-a (TNF-a), soluble TNF-receptors I and II (sTNF-RI, sTNF-RII), interleukin-6
(IL-6) and interferon-gamma (IFN-y). Controlled for age and race, high CRP
concentration was associated with increased BPH risk (Odds Ratio (OR) =1.40; 95%
confidence interval (CI): 1.04, 1.88 for Q4 vs. Ql); this was attenuated after control for
body mass index (OR=l .30; 95%CI: 0.95, 1.75). Low sTNF-RII and high IL-6
concentrations were associated with increased BPH risk (sTNF-RII: OR=0.61; 95%CI:
0.46, 0.82; IL-6: OR=1.79; 95%CI: 1.32, 2.42, for Q4 vs. Ql); these associations were
in only men<65 years old. Results suggest that systemic inflammation or lower levels
of soluble receptors that bind inflammatory cytokines increase BPH risk.
Introduction
Benign prostatic hyperplasia (BPH) is a common medical conditions among
43
middle-aged and older males, affecting 40 to 50 percent of men by age 50 and nearly
80 percent of men by age 70[3,104]. Histologically, BPH is characterized by hyper-
proliferation of the stromal and, to a lesser extent, epithelial regions of the prostate,
which causes a constellation of lower urinary tract symptoms (LUTS) that affect
quality of life and prompt many men to seek treatment[l,5,8]. Medical and surgical
treatment for BPH symptoms is expensive[l 14], and as the US population ages, both
the number of men affected and the associated medical costs are expected to rise[l 15].
The pathogenesis of BPH remains poorly understood; however, it is likely that
inflammation plays a role in its development or progression. Histological evidence of
acute and chronic inflammation is commonly found in prostate biopsy and BPH
specimens[62,64], and intra-prostatic inflammation is associated with several
characteristics of BPH[63,66,68]. Furthermore, inflammatory cytokines, which
function as potent mitogens capable of inducing the hyperplastic changes characteristic
of BPH[74], are over-expressed in BPH tissue[70,72].
The underlying causes of intra-prostatic inflammation remain unclear, though
several hypotheses have been proposed including: response to tissue damage caused by
infection[76] or other causes[77,78], or an autoimmune response[79]. It is also
plausible that systemic inflammation could contribute to the initiation or progression
within the prostate. Obesity and abdominal obesity, which are characterized by large
deposits of adipose tissue that produce an excess of inflammatory cytokines[26], are
well-established risk factors for BPH[12,13,54]. Thus, it is reasonable to hypothesize
44
that adipose-derived increases in circulating cytokine concentrations may influence
BPH risk.
This report examines whether systemic concentrations of several inflammatory
markers elevated in obesity, including tumor necrosis factor-alpha monomer (TNF-a),
TNF receptors I and II (sTNF-RI and II), interleukin-6 (IL-6), interferon-gamma (IFN-
y), and C -reactive protein (CRP) affect the risk of symptomatic BPH, and whether
these factors mediate the relationship between obesity and BPH risk.
Methods
Data for this study have been described in detail previously [105]. Briefly, data
are from the Prostate Cancer Prevention Trial (PCPT), a randomized, placebo-
controlled trial testing whether finasteride reduced prostate cancer risk. [92] Analyses
are restricted to the 9,457 placebo-arm participants. Exclusion criteria for these
analyses included self-reported history of BPH (n= 1,904), medical or surgical
treatment for BPH (n=701), IPSS>7 (n=l,820) at baseline, and use of steroid hormones
(n=61), leaving 4,971 men eligible for this study.
Extensive medical data was collected at baseline, 6-month, and annual clinic
visits and at every 3- and 9-month phone contact between visits. At recruitment (3
months prior to baseline), baseline, and annual clinic visits, participants completed the
IPSS[6], a 7-item self-administered questionnaire assessing the frequency of LUTS. At
baseline, clinic staff measured height and weight. Age, race/ethnicity, physical activity,
alcohol consumption, and history of smoking were collected at using self-administered
45
questionnaires.
Definition of BPH Cases and Controls
Incident BPH was defined either as a report of treatment for or development of
significant LUTS. Treatments included use of ot-blockers, finasteride or surgical
intervention. Development of significant symptoms was defined as either (1) two IPSS
scores >14; or (2) substantial increase in LUTS from baseline (two IPSS scores at least
5 units higher than baseline plus at least one score >12). The latter is a more
conservative version of the definition of BPH progression used in the Medical Therapy
of Prostatic Symptoms Trial[l 16], in which a single increase of 4 defined significant,
clinical progression. There were a total of 727 incident BPH cases which were further
classified by type of BPH event (treatment or symptoms) and by predominant type of
BPH symptom (mostly obstructive, mostly irritative or mixed obstructive/irritative).
We defined type of BPH symptom using the IPSS at or prior to (for treatment) the
defining BPH event; obstructive (incomplete emptying, interrupted stream, weak
stream, straining) and irritative symptoms (frequency, urgency, nocturia) IPSS items
were weighted to contribute 50% to total IPSS. If the obstructive or irritative score
contributed >55% to total IPSS, BPH type was categorized by that symptom, otherwise
BPH type was categorized as mixed obstructive/irritative.
Controls were drawn from the 1,497 men who, during the 7-year trial, had no
more than 2 missing IPSS scores, no single IPSS greater than 7, and reported no
diagnosis of BPH or treatment for BPH. From this sample, we selected all men aged
46
70 years and over and all non-Caucasian men, to maximize statistical power when
examining these subgroups. Remaining controls were randomly selected to yield a
total sample (n = 727) frequency matched to the age distribution (in 5 year age groups)
of cases.
Blood Collection, Processing and Laboratory Analysis
Bloods were drawn at recruitment into a 7 ml EDTA vacutainer tube and
shipped overnight to a storage facility, where they were centrifuged, aliquoted and
stored at -70° C. These analyses used serum that had gone through one freeze-thaw
cycle.
Serum concentrations of CRP and cytokines were quantified using
SearchLight® multiplex enzyme-linked immunosorbent assay (Pierce Biotechnology
Inc., Woburn, MA). Each array consisted of a 96-well plate pre-spotted with target
antibodies for the specified analytes. Plates were imaged using SearchLight® Black
Ice Imaging system and analyzed using SearchLight® Array Analyst software.
Serum was not available from 39 men, and 56 men with evidence of active
infection (CRP>10mg/L) were excluded, leaving data from 676 cases and 683 controls
available for these analyses. The lower limits of detection (LLD) and mean
coefficients of variation (derived from samples run in triplicate at various dilutions)
were: CRP: 5.86pg/mL, 8.0%; TNF-a: 2.34pg/mL, 15.3%; sTNF-Rl:0.78pg/mL,
8.9%; sTNF-RII: 0.39pg/mL, 8.3%; IL-6:0.39pg/mL, 11.2%, and IFN-y: 0.39pg/mL,
15.2%, respectively. For observations with a detectable concentration below the LLD
47
(TNF-a: n=322; IL-6: n=32; and IFN-y: n=225), the detected concentration was used.
For observations with undetectable concentrations, a value halfway between 0 and the
lowest detectable concentration was assigned, which equaled 0.8 for TNF-a (n=40),
0.1/mL for IL-6 (n=26), and O.lpg/mL for IFN- y (n=10).
Statistical Methods
Descriptive statistics were used to characterize the study sample and
distribution of CRP and cytokine concentrations in cases and controls. Serum
concentrations of analytes were categorized into quartiles using the distribution in
controls, and unconditional logistic regression was used to calculate odds ratios (OR)
and 95% Confidence Intervals (CI) for BPH risk. All models were adjusted for
matching variables (age at baseline (continuous), race (white, other)), and additional
models included BMI (continuous) and IPSS at baseline (continuous). Controlling for
other factors associated with BPH risk in this sample, including waist to hip ratio[12],
insulin-like growth factors (IGF-1, and IGFBP-3)[105], steroid hormones (testosterone,
estradiol and 3-a-diol glucaronide)[106], alcohol consumption and diet (fat, protein,
red meat, and vegetables)[107], and controlling for medical conditions strongly
associated with either BPH risk and/or systemic inflammation (diabetes, arthritis and
cardiovascular disease) did not affect risk estimates. Tests for linear trend were
performed using an ordinal variable corresponding to rank (lowest to highest
quartile)[108].
48
Additional analyses were completed stratified by age, BMI, physical activity,
and smoking status. Interaction tests were based on P-values for the interaction term of
CRP or cytokine concentration rank (lowest to highest tertile) times a dummy variable
for age (<65 vs. >65), physical activity (sedentary/light vs. moderate/very active),
smoking status (current vs. former/never), or BMI (<25 vs. >25). Polytomous logistic
regression models were used to calculate separate ORs for outcomes defined by time
from baseline to BPH event (<4 or >4 years), type of BPH event (BPH treatment or
symptoms), and type of BPH symptom (irritative, mixed obstructive/irritative, or
obstructive). We also considered whether the previous finding of an increased BPH
risk associated with obesity[12] was mediated by CRP or cytokines, by examining the
association of obesity with BPH risk unadjusted and adjusted for analytes. All P-
values were two-sided and considered statistically significant at .P<0.05. Statistical
analyses were conducted using SAS (version 9.1 Cary, NC).
Results
Distributions of participant characteristics are given in Table 2.1. Participants
were mostly white, overweight or obese, and nonsmokers. Compared to controls, men
with BPH consumed less alcohol, were more likely to have cardiovascular disease and
had a higher IPSS at baseline. Table 2.2 gives age and race adjusted geometric mean
and median of CRP and cytokine concentrations in cases and controls. Differences in
geometric mean concentrations between cases and controls were statistically significant
for sTNF-Rl (+5.5%), sTNF-RII (-8.8%), IL-6 (+24.4%), and IFN-y (+13.9%), and
49
borderline significant for CRP (+10.2%).
Table 2.3 gives adjusted odds ratios for risk of BPH associated with serum CRP
and cytokine concentrations. In Model 1 (adjusted for age and race), men in the
highest quartile of CRP had a significantly increased risk of symptomatic BPH
(OR=1.40; 95%CI: 1.04, 1.88; P
tre
nd =0.03); however, after adjustment for BMI the
association was attenuated and no longer statistically significant. In Model 2 (adjusted
for matching variables and BMI), men with sTNFR-II concentrations above the lowest
quartile had a significantly reduced risk of BPH, with no evidence of a linear dose-
response association. In a post-hoc analysis contrasting quartile 1 with quartiles 2-4,
high sTNFR-II was associated with a 46% (95% CI: 32%, 58%; p<0.0001) lower BPH
risk. Men in the highest two quartiles of IL-6 had a 46% and 79% higher risk of BPH,
respectively, with evidence of a dose-response association (P
t
rend
<
0.001). There was a
suggestion of increased BPH risk in the highest quartile for sTNF-RI and IFN- y, which
is consistent with the significant difference in geometric means between cases and
controls (Table 2.2); however, the trend test did not reach statistical significance. There
was no association of TNF-a with BPH risk. To control for pre-clinical disease,
models were also adjusted for IPSS at baseline (Model 3); however, results were
similar to fully-adjusted models (Model 2). Results were similar when analyses
excluded men diagnosed with prostate cancer (n=93 cases, n=105 controls; Table 2.4),
or who died during the PCPT (n=22 cases, n=12 controls; Table 2.5).
Associations of CRP and cytokines differed little across strata defined by BMI
50
(Table 2.6), smoking status (Table 2.7), and physical activity (Table 2.8); although
there was an increased risk of BPH for the highest concentration IFN- y among normal
weight men only. However, for strata defined by age, there were strong associations
between sTNF-RII and IL-6 and BPH risk among younger men only (Pinteraction
=:
0.02
and .Pinteraction
=
0.03, respectively; Table 2.9), and no associations among older men.
There were no substantial differences in associations of CRP, cytokines and BPH risk
when stratified by type of BPH event (Table 2.10) and time between blood collection
and BPH event (Table 2.11).
To further investigate the difference in associations between CRP, cytokines
and BPH risk between younger and older men, we examined whether the type of LUTS
differed by age and whether the associations for CRP, sTNF-RII and IL-6 and BPH
risk among younger men differed by type of LUTS. The distributions of LUTS type
were similar by age group: the proportion of men reporting mostly irritative, mixed, or
mostly obstructive symptoms, respectively, was 57.7%, 18.0%, 24.3% for 55-59 year
olds; 48.5%, 28.4%, 23.0% for 60-64 year olds; 59.2%, 22.0%, 18.9% for 65-69 year
olds; and 53.7%, 25.9%, 20.4% for men over 70 years (P=0.31). Furthermore,
associations between CRP, sTNF-RII and IL-6 and BPH, either in the total population
(Table 2.12) or among men less than 65 years at baseline Table 2.13), did not differ by
type of LUTS.
Obesity was associated with an increased risk of BPH in this cohort[12];
therefore, additional analyses examined whether CRP or cytokines mediate the
51
association between obesity and risk of BPH. Compared to men who were normal
weight, overweight and obese men had a 26% (OR=1.24, 95%CI 0.97-1.63) and 37%
(OR=1.37,95%CI 1.00-1.87) increased risk of BPH, respectively (P
frend
=0.05).
Control for IL-6 decreased the odds ratio between obesity and BPH risk by 35%, from
1.37 to 1.24 (95%CI, 0.90-1.71; P
tre
nd=0.17) while control for sTNF-RII increased the
odds ratio between obesity and BPH risk by 11%, from 1.37 to 1.41 (95%CI, 1.03-
1.92; P
t
rend
=
0.03). In models with CRP or obesity alone, both were significantly and
positively associated with BPH risk (CRP Q4 vs. Ql OR=1.40; 95%CI: 1.04-1.88;
^trend=0.03; obese vs. normal weight OR=1.37; 95%CI: 1.00-1.87; / Wr O. 04) .
However, in a model with both CRP and obesity, associations for both factors were
attenuated and neither reached statistical significance (CRP Q4 vs. Ql OR=1.30;
95%CI: 0.95-1.77; P,
rend
=0.15; obese vs. normal weight OR=1.28; 95%CI: 0.92-1.77;
-Ptrend=0.12). In models controlled for one or more cytokines, once IL-6 was added to
the model, control for CRP and/or sTNF-RII had no further effect on the association of
obesity and BPH risk (Table 2.14).
Discussion
In this prospective study, circulating levels of inflammatory markers were
associated with risk of incident, symptomatic BPH. Specifically, high serum CRP
concentrations (Q4) were associated with an increased risk of BPH with no dose-
response across lower quartiles (Q1-Q3), low serum sTNF-RII concentrations (Ql)
were associated with a significantly increased risk of BPH with no dose-response
52
across higher quartiles (Q2-Q4), and high serum IL-6 concentrations were associated
with a dose-response increased risk of BPH. These associations did not differ by BMI,
smoking status, physical activity, type of BPH event, or time between blood draw and
BPH event; however, associations of sTNF-RII and IL-6 with BPH risk were limited to
men <65 years at baseline. There was also a suggestion of increased risk in the highest
quartile for sTNF-Rl and IFN-y; however, the associations were not statistically
significant.
Our findings of an association between markers of systemic inflammation and
BPH risk are supported by multiple lines of evidence suggesting that inflammation
plays a role in the etiology of BPH. Histologic studies have found acute and/or chronic
inflammation in up to 100% of BPH specimens [62-65]. In cross-sectional studies, the
presence of inflammatory infiltrates in prostate tissue is associated with several
measures of BPH including increased prostate volume[62,63], more severe LUTS[66],
acute urinary retention (AUR)[67,68], and epithelial cell proliferation[69].
Furthermore, in-situ studies have found that expression of pro-inflammatory cytokines
is increased in BPH tissue[70-73], and in a rat prostate model, administration of
immunostimulatory compounds induces epithelial proliferation and hyperplastic
lesions similar to BPH nodules[74]. In recent prospective studies, participants with
acute inflammation in biopsy specimens had a greater risk of BPH progression (LUTS)
and AUR[63]. Furthermore, in a large population-based cohort of men, daily users of
non-steroidal anti-inflammatory drugs (NSAIDS) had a lower risk of several clinical
53
measures of BPH (low maximum flow rate, increased prostate volume and elevated
PSA), and a lower risk of developing moderate/severe LUTS[75].
Only three previous studies have investigated the association between systemic
inflammatory markers and BPH risk [80,81,117]. Our findings are consistent with one
small, hospital-based case-control study in which men with histologically confirmed
BPH had higher concentrations of IL-6 than controls (1.9 vs.0.7pg/ml,
respectively)[80]. Our results are also consistent with a cross-sectional study which
found that men with elevated CRP concentrations had a non-significant increased risk
of more severe LUTS.[81]. In a small prospective study, high CRP concentrations
were not associated with an absolute increase in LUTS, although men with high CRP
concentrations were more likely to have a rapid proportional increase in irritative
LUTS[117].
Soluble TNF-Rs reflect TNF-a activity because they are shed from cellular
TNF-Rs in response to TNF-a binding; therefore, we originally hypothesized that high
sTNF-R concentrations would be associated with increased risk of BPH. However,
sTNF-Rs also have direct anti-inflammatory affects by either down-regulating the
expression of cell membrane TNF-Rs and decreasing the sensitivity of target cells to
TNF-a[ 118], or by binding to, and thereby acting as a competitive antagonist for TNF-
a[l 19]. Thus, the interpretation of low sTNF-R concentrations is complex, and in this
study may reflect an inability to modulate responses to TNF-a.
The associations between sTNF-RII and IL-6 and BPH risk in young men only
54
are difficult to understand. Some studies are consistent with our findings: one reported
a much stronger association between CRP and BPH risk among younger men[81], and
another found an association between the serologic presence of sexually transmitted
infections and symptomatic BPH among men aged30-59 but not among men >60
[120]. In secondary analyses we found no age-related differences in the type of BPH
symptoms reported, and associations between cytokines and BPH risk did not differ by
symptom type. It is possible that BPH etiology differs in younger and older men;
however, this needs confirmation in further studies.
Both in this nested case-control sample and the PCPT overall, there were
significant associations of obesity with increased BPH risk[12]. However, when
examining whether this association was mediated by CRP concentrations, both CRP
and BMI were attenuated to a similar degree and neither reached statistical
significance. This finding suggests that BMI and CRP jointly contribute to BPH risk
and the effects of one cannot be separated from the other, although we cannot rule out
the possibility that obesity and CRP are either markers of, or confounded by, another
risk factor for BPH. Furthermore, when examining whether IL-6 concentrations
mediate the association between obesity and BPH, adjustment for IL-6 resulted in only
a moderate attenuation (35%) of the association, suggesting that systemic inflammation
is not the only mechanism through which obesity affects BPH risk. However, it is also
possible that measurement error in IL-6 limited the ability to explain the association
between obesity and BPH.
55
We found no associations between TNF-a, sTNF-Rl, IFN-y and BPH risk. An
increased risk of BPH was found for the highest concentration IFN- y among normal
weight men; however, this is inconsistent with the other cytokines and may be a chance
finding. The lack of associations may be due to low cytokine concentrations,
particularly for TNF-a and IFN-y, where 27% and 17% of samples were below the
limits of detection. sTNF-RI is derived from epithelial cells whereas sTNF-RII
originates primarily from activated T-cells, B cells, and neutrophils.[121] It is possible
that activation of the TNF-Rs results in distinct activities [122] and that TNF-RII plays
a stronger role in the inflammatory response.
One important strength of this study is the prospective design. Men with a
history of BPH or BPH symptoms at the time of blood draw were excluded and all
cases were incident. Cross-sectional and case-control studies of inflammation and
BPH are problematic because differences in serum cytokine concentrations could be
due to BPH. In addition, this study used a rigorous definition of BPH, which captured
all current medical treatments as well as substantial LUTS, and specifically excluded
men who reported transient elevations in IPSS or a physician diagnosis of BPH in the
absence of symptoms or treatment. This is in contrast to previous studies which relied
on a single incomplete assessment of LUTS[81], or surgery alone as an endpoint[80].
There are several important limitations to this study. Our definition of BPH by
LUTS is not specific and cannot distinguish between LUTS due to BPH and LUTS due
to other urologic, neurologic or bladder conditions[l 12]. However, there are two key
56
arguments to suggest that our results are not substantially affected by LUTS unrelated
to BPH. First, we estimate that the prevalence of 'other conditions' associated with
LUTS in the PCPT study population is only 6.5% (based on the age-specific
prevalences of these conditions (13.5%) within the Olmstead County cohort which also
includes men with a history of prostate surgery)[l 13]. Second, there were no
differences in associations between cytokines and BPH by symptom type (mostly
obstructive vs. mostly irritative), suggesting that the observed association was not due
to prostate enlargement alone. We deliberately selected a control group free of
significant LUTS throughout the 7-year trial; thus, men who developed mild to
moderate symptoms were excluded from this study. It is unclear whether the
development of mild/moderate symptoms is indicative of an intermediate form of BPH,
but if so, inclusion of these men in the controls would have attenuated our results. We
could not control for use of NSAIDs because it was not well captured in the PCPT.
One study reported modest inverse associations between NSAIDs and symptomatic
BPH [75], and thus it is possible that NSAIDS somewhat confound our findings.
Lastly, CRP and cytokine concentrations are subject to non-differential measurement
error. CRP and cytokine concentration from a single baseline blood specimen may not
accurately reflect concentration at the physiologically relevant time-point, and bloods
were non-fasting and drawn at all times during the day, which may have increased
variability in concentrations. .
This study supports the hypothesis that systemic inflammation plays a role in
57
the development or progression of symptomatic BPH. High CRP concentrations were
associated with a decreased risk of BPH, although the association was attenuated after
adjustment for BMI, suggesting that BMI and CRP jointly contribute to BPH risk.
Low sTNF-RII and high IL-6 concentrations were associated with an increased risk of
BPH; however, only among men younger than 65 years, suggesting that the
pathogenesis of BPH may differ between younger and older men. CRP and IL-6
explained only a moderate proportion (19-35%) of the association between obesity and
increased BPH risk; thus, elevations in these inflammatory markers may not be the
only mechanism through which obesity affects risk of BPH. Several other factors
including chronic infection or autoimmune inflammatory disorders could also
contribute systemic inflammation and to the development of BPH. Future studies will
be needed to confirm our findings and assess the role of other sources of systemic
inflammation.
58
Table 2.1. Demographic and lifestyle characteristics of the study population '
Age
55-60
60-65
65-70
70+
Race/Ethnicity
White
Other
Waist to hip ratio
d
<0.95
0.95-0.99
1.00-1.04
1.05+
Body Mass Index
c
'
d
Normal (<25)
Overweight (25-29.9)
Obese (>30)
Alcohol Consumption
d
< 1 drink/month
1-3 drinks/month
1 - 6 drinks/week
7- 13 drinks/week
> 14 drinks/week
Physical activity
d
Sedentary
Light activity
Moderate activity
Very active
Current Smoker
d
Prevalent Cardiovascular
Disease
e
Prevalent Arthritis
e
Prevalent Diabetes
e
Baseline IPSS
0-3
4-5
6-7
Benign Prostatic Hyperplasia
Cases (N=676)
n
205
188
179
104
626
50
251
218
119
18
161
350
154
191
93
230
97
64
108
304
192
69
50
119
28
5
296
278
102
(%)
(30.3)
(27.8)
(26.5)
(15.4)
(92.6)
(7.4)
(41.4)
(36.0)
(19.6)
(3.0)
(24.2)
(52.6)
(23.2)
(28.3)
(13.8)
(34.1)
(14.4)
(9.5)
(16.1)
(45.2)
(28.5)
(10.3)
(7.4)
(17.6)
(4.1)
(0.7)
(43.8)
(41.2)
(15.1)
Controls (N=683)
n
207
188
179
109
638
45
294
220
111
19
194
346
140
151
121
28
107
76
96
286
228
71
41
74
21
1
531
137
15
(%)
(30.3)
(27.5)
(26.2)
(16.0)
(93.4)
(7.4)
(45.7)
(34.2)
(17.2)
(3.0)
(28.5)
(50.9)
(20.6)
(22.1)
(17.7)
(33.4)
(15.7)
(11.1)
(14.1)
(42.0)
(33.5)
(10.4)
(6.0)
(10.8)
(3.1)
(0.2)
(77.8)
(20.1)
(2.2)
(BPH)
P"
0.99
0.56
0.47
0.17
0.04
0.23
0.57
0.0004
0.29
0.10
O. 01
a
Excludes 23 cases and 16 controls with missing cytokine data, and 28 cases and 28 controls with CRP>10mg/L at
baseline.
b
p from chi-square test
"Weight (kg)/height (m)
2
"# participants missing data for WHR: 109, BMI: 11 cases and 3 controls, alcohol: 1, physical activity:5, smoking
status:2
'Assessed at baseline
59
Table 2.2. Serum concentrations of cytokines for cases and controls
3
C-reactive protein, (mg/L)
Median
Geometric mean
b
(95%CI)
5
th
- 95
th
percentiles
TNF-a monomer, (pg/mL)
Median
Geometric mean
b
(95%CI)
5
th
- 95
th
percentiles
sTNF-Rl, (pg/mL)
Median
Geometric mean
b
(95%CI)
5
th
- 95
th
percentiles
sTNF-RII, (pg/mL)
Median
Geometric mean
b
(95%CI)
5
th
- 95
th
percentiles
IL-6, (pg/mL)
Median
Geometric mean
b
(95%CI)
5
th
- 95
th
percentiles
IFN-y, (pg/mL)
Median
Geometric mean
b
(95%CI)
5
th
- 95
th
percentiles
Benign Prostatic Hyperplasia
Cases (N=676)
1.49
1.46(1.31,1.63)
0.26 - 6.20
10.6
10.8 (9.2, 12.7)
0.8-107.2
1,887
1841 (1757, 1929)
1,035-3,770
850
820 (780, 820)
475-1, 620
3.30
3.19(2.86,3.55)
0.6-11.0
2.40
2.58(2.26,2.94)
0.4-14.4
Controls (N=683)
1.32
133(1.19,1.49)
0.27-5.55
10.6
11.2(9.5,13.1)
0.3-100.4
1,825
1746(1665,1830)
39-1, 770
960
899 (855, 945)
390-1, 770
2.80
2.56(2.30,2.86)
0.4-7.8
2.20
2.27(1.99,2.59)
0.4-12.4
(BPH)
P
0.04
c
0.07
d
0.67
c
0.65
d
0.08
c
0.01
d
<0.0001
c
<0.0001
d
O. 0001
c
O.0001
d
0.02
c
0.03
d
"Excludes 23 cases and 16 controls with missing cytokine data, and 28 cases and 28 controls with CRP>10mg/dL at
baseline.
b
Adjusted for age and race
c
Kruskall-Wallis test OR Wilcoxon rank-sum
d
T-test
60
Table 2.3. Main effects of CRP and cytokines on Benign Prostatic Hyperplasia risk
2
No. cases/
controls
Model 1
b
Model 2
c
Model 3
d
OR (95% CI) OR (95% CI) OR (95% CI)
CRP, (mg/L)
Ql 151/171 1.00 1.00 1.00
Q2 159/171 1.05 (0.77,1.43) 1.04 (0.76, 1.41) 1.14 (0.81, 1.60)
Q3 155/171 1.02 (0.75,1.39) 0.98 (0.72, 1.34) 1.01 (0.72, 1.43)
Q4 211/170 1.40 (1.04,1.88) 1.30 (0.95,1.78) 1.38 (0.98,1.94)
P for trend 0.03 0.12 0.11
TNF-a, (pg/mL)
Ql 170/171 1.00 1.00 1.00
Q2 169/171 0.99 (0.73, 1.34) 0.99 (0.73, 1.34) 0.97 (0.70, 1.35)
Q3 178/172 1.03 (0.76, 1.38) 1.03 (0.76, 1.40) 1.13 (0.81, 1.57)
Q4 159/169 0.94 (0.70, 1.28) 0.96 (0.71, 1.31) 1.07 (0.77, 1.50)
P for trend 0.73 0.88 0.51
sTNF-RI, (pg/mL)
Ql 170/172 1.00 1.00 1.00
Q2 143/170 0.85 (0.62,1.15) 0.84 (0.62, 1.15) 0.89 (0.63, 1.24)
Q3 158/171 0.93 (0.69, 1.26) 0.92 (0.68, 1.25) 0.88 (0.63, 1.23)
Q4 205/170 1.22 (0.90,1.63) 1.21 (0.90, 1.63) 1.30 (0.94, 1.80)
P for trend 0.14 0.16 0.12
sTNF-RII, (pg/mL)
Ql 262/174 1.00 1.00 1.00
Q2 140/168 0.55 (0.41,0.74) 0.55 (0.41,0.74) 0.56 (0.40,0.77)
Q3 117/174 0.45 (0.33,0.61) 0.45 (0.33,0.61) 0.50 (0.36,0.69)
Q4 157/167 0.62 (0.46,0.83) 0.61 (0.46,0.82) 0.67 (0.49,0.92)
P for trend <0.001 <0.01 <0.01
IL-6, (pg/mL)
, Ql 142/192 1.00 1.00 1.00
Q2 105/146 0.97 (0.70, 1.36) 0.96 (0.69, 1.35) 0.92 (0.64, 1.33)
Q3 197/178 1.50 (1.10,2.02) 1.46 (1.08,1.97) 1.28 (0.92,1.78)
Q4 232/167 1.88 (1.40,2.53) 1.79 (1.32,2.42) 1.37 (1.20,2.32)
P for trend <0.001 O.001 <0.001
IFN-y, (pg/mL)
Ql 180/189 1.00 1.00 1.00
Q2 140/169 0.87 (0.64, 1.18) 0.87 (0.64, 1.18) 0.85 (0.61, 1.19)
Q3 164/161 1.07 (0.79, 1.44) 1.06 (0.79, 1.43) 1.10 (0.79, 1.53)
Q4 192/164 1.22 (0.91,1.64) 1.24 (0.92,1.66) 1.31 (0.95,1.81)
P for trend (HO (^09 (X05
a
Quartile outpoints: CRP: 0.67, 1.32,2.48; TNF-a: 4.5, 10.6, 24.4TNF-R1: 1490, 1825, 2235; TNF-R2: 740, 960, 1185; IL-
6: 1.6,2.8, 4.2; IFN-y: 1.2,2.2,4.0.
b
Adjusted for matching covariates only (age at baseline and race)
c
Adjusted for matching covariates, body mass index(linear)
d
Adjusted for matching covariates, body mass index(linear), and IPSS at baseline
61
Table 2.4. Main effects of CRP and cytokines on
Benign Prostatic Hyperplasia risk, excluding men
with prostate cancer
3
CRP, (mg/L)
Qi
Q2
Q3
Q4
P for trend
No. cases/
controls
126/138
132/146
137/146
187/148
TNF-a, (monomer)
Qi
Q2
Q3
Q4
P for trend
142/142
149/142
156/148
135/146
sTNF-RI, (pg/mL)
Qi
Q2
Q3
Q4
P for trend
146/148
119/140
137/144
180/146
sTNF-RII, (pg/mL)
Qi
Q2
Q3
Q4
P for trend
IL-6, (pg/mL)
Qi
Q2
Q3
Q4
P for trend
IFN-y, (pg/mL)
Qi
Q2
Q3
Q4
P for trend
222/151
119/149
101/144
140/134
118/158
88/123
170/153
206/144
158/151
119/144
139/142
166/141
OR
1.00
1.05
1.02
1.40
1.00
0.99
1.03
0.94
1.00
0.85
0.93
1.22
1.00
0.55
0.45
0.62
1.00
0.97
1.50
1.88
1.00
0.87
1.07
1.22
Model 2
b
(95% CI)
(0.77,1.43)
(0.75, 1.39)
(1.04, 1.88)
0.03
(0.73, 1.34)
(0.76, 1.38)
(0.70, 1.28)
0.73
(0.62, 1.15)
(0.69, 1.26)
(0.90, 1.63)
0.14
(0.41,0.74)
(0.33,0.61)
(0.46, 0.83)
0.01
(0.70, 1.36)
(1.10,2.02)
(1.40,2.53)
O.001
(0.64,1.18)
(0.79, 1.44)
(0.91,1.64)
0.34
a
Excludes 93 cases and 105 controls with prostate cancer
b
Adjusted for matching covariates only (age at baseline
and race)
62
Table 2.5. Main effects of CRP and cytokines on
Benign Prostatic Hyperplasia risk, excluding men who
died during the PCPT"
CRP, (mg/L)
Qi
Q2
Q3
Q4
P for trend
No. cases/
controls
148/168
153/168
148/168
205/167
TNF-a, (monomer)
Qi
Q2
Q3
Q4
P for trend
165/168
162/169
171/168
156/166
sTNF-RI, (pg/mL)
Qi
Q2
Q3
Q4
P for trend
166/171
139/167
152/166
197/167
sTNF-RII, (pg/mL)
Qi
Q2
Q3
Q4
P for trend
IL-6, (pg/mL)
Qi
Q2
Q3
Q4
P for trend
IFN-Y, (pg/mL)
Qi
Q2
Q3
Q4
P for trend
254/171
136/167
110/167
154/166
138/190
102/144
191/176
223/161
174/189
138/166
159/156
183/160
OR
1.00
1.05
1.02
1.40
1.00
0.99
1.03
0.94
1.00
0.85
0.93
1.22
1.00
0.55
0.45
0.62
1.00
0.97
1.50
1.88
1.00
0.87
1.07
1.22
Model 2
b
(95% CI)
(0.77, 1.43)
(0.75, 1.39)
(1.04, 1.88)
0.03
(0.73, 1.34)
(0.76, 1.38)
(0.70, 1.28)
0.73
(0.62,1.15)
(0.69, 1.26)
(0.90, 1.63)
0.14
(0.41,0.74)
(0.33, 0.61)
(0.46, 0.83)
0.01
(0.70, 1.36)
(1.10,2.02)
(1.40,2.53)
O.001
(0.64,1.18)
(0.79, 1.44)
(0.91,1.64)
0.34
a
Excludes 93 cases and 105 controls with prostate cancer
Adjusted for matching covariates only (age at baseline
and race)
63
Table 2.6. Associations of CRP and cytokines with Benign Prostatic Hyperplasia risk,
stratified by body mass index
3
CRP, (mg/L)
Tl
T2
T3
TNF-a, (pg/mL)
Tl
T2
T3
No. cases/
controls
70/92
48/61
43/41
57/65
61/65
43/64
sTNF-RI, (pg/mL)
Tl
T2
T3
52/65
42/65
67/64
sTNF-RII, (pg/mL)
Tl
T2
T3
IL-6, (pg/mL)
Tl
T2
T3
IFN-Y, (pg/mL)
Tl
T2
T3
79/67
33/64
49/63
53/87
41/56
67/51
39/56
49/77
73/61
BMK25
OR
1.00
0.84
1.29
1.00
1.07
0.74
1.00
0.79
1.23
1.00
0.41
0.61
1.00
1.27
2.34
1.00
0.89
1.82
(95% CI)
(0.43, 1.61)
(0.61,2.72)
P =
L
interaction
(0.64, 1.76)
(0.43, 1.26)
P . • —
•* interaction
(0.46, 1.36)
(0.74,2.10)
P • • =
L
interaction
(0.24,0.71)
(0.37, 1.02)
P =
J
interaction
(0.74,2.17)
(1.38,2.95)
P =
L
interaction
(0.51,1.54)
(1.06,3.13)
P • =
1
interaction
No. cases/
controls
135/135
151/166
218/185
0.66
173/166
166/161
165/160
0.39
161/162
159/164
184/160
0.66
230/165
126/158
148/164
0.99
121/155
167/150
216/181
0.22
155/162
169/163
180/161
0.17
BMI25+
OR
1.00
1.03
1.18
1.00
0.99
1.00
1.00
0.98
1.16
1.00
0.58
0.64
1.00
1.43
1.49
1.00
1.08
1.15
(95% CI)
(0.59, 1.81)
(0.67, 2.08)
(0.73, 1.34)
(0.73, 1.35)
(0.72, 1.34)
(0.85, 1.57)
(0.42, 0.79)
(0.47, 0.87)
(1.03, 1.98)
(1.09,2.04)
(0.80, 1.48)
(0.85, 1.58)
"Adjusted for matching covariates, and body mass index(linear)
64
Table 2.7. Associations of CRP and cytokines with Benign Prostatic Hyperplasia risk,
stratified by smoking status
3
CRP, (mg/L)
Tl
T2
T3
TNF-a, (pg/mL)
Tl
T2
T3
Former/Never Smoker
No. cases/
controls
184/212
193/214
238/212
214/218
213/211
198/212
sTNF-RI, (pg/mL)
Tl
T2
T3
205/218
184/210
236/213
sTNF-RII, (pg/mL)
Tl
T2
T3
IL-6, (pg/mL)
Tl
T2
T3
IFN-Y, (pg/mL)
Tl
T2
T3
285/217
152/212
178/209
170/245
203/202
242/191
169/183
221/247
235/211
OR
1.00
1.00
1.20
1.00
1.03
0.96
1.00
0.91
1.15
1.00
0.59
0.65
1.00
1.65
1.83
1.00
0.96
1.20
(95% CI)
(0.76, 1.33)
(0.90, 1.59)
P • • =
±
interaction
(0.79, 1.35)
(0.73,1.26)
P . =
* interaction
(0.69, 1.20)
(0.88,1.51)
P. . =
L
interaction
(0.45, 0.78)
(0.50, 0.85)
P =
•* interaction
(1.27,2.16)
(1.38,2.43)
P =
•* interaction
(0.73, 1.27)
(0.90, 1.59)
P =
1
interaction
Current Smoker
No. cases/
controls
17/14
12/14
20/13
0.66
21/14
17/14
12/13
0.43
12/14
13/14
25/14
0.17
22/12
11/14
16/15
0.19
13/8
15/14
21/19
0.69
17/14
10/11
23/16
0.92
OR
1.00
0.54
1.05
1.00
0.85
0.69
1.00
0.91
2.75
1.00
0.22
0.58
1.00
0.40
1.15
1.00
0.93
1.70
(95% CI)
(0.17, 1.69)
(0.35,3.10)
(0.29, 2.48)
(0.22,2.14)
(0.28, 2.93)
(0.87, 8.72)
(0.07, 0.69)
(0.17,1.98)
(0.13, 1.20)
(0.36, 3.64)
(0.29, 3.05)
(0.58, 5.02)
'Adjusted for matching covariates, and body mass index(linear)
65
Table 2.8. Associations of CRP and cytokines with Benign Prostatic Hyperplasia risk,
stratified by physical activity
3
CRP, (mg/L)
Tl
T2
T3
TNF-a, (pg/mL)
Tl
T2
T3
Sedentary/Light Activity
No. cases/
controls OR (95% CI)
123/127
121/126
161/127
133/130
142/125
137/127
sTNF-RI, (pg/mL)
Tl 120/129
T2
T3
118/125
174/128
sTNF-RII, (pg/mL)
Tl 185/129
T2
T3
IL-6, (pg/mL)
Tl
T2
T3
IFN-Y, (pg/mL)
Tl
T2
T3
89/125
131/126
101/131
159/131
145/118
107/106
15/152
150/124
1.00
0.96
1.22
1.00
1.12
1.06
1.00
1.00
1.43
1.00
0.50
0.70
1.00
1.54
1.52
1.00
0.99
1.18
0.67, 1.37
0.86,1.74
P • =
•* interaction
(0.79, 1.58)
0.75,1.49
P - =
•* interaction
0.70, 1.43
1.02,2.03
P • • -
x
interaction
0.35,0.71
0.50,0.98
P • =
•* interaction
1.08,2.18
1.06,2.20
P • =
* interaction
0.70, 1.41
0.82, 1.70
P =
-* interaction
Moderate/Very Active
No. cases/
controls OR (95% CI)
70/100
92/99
95/99
0.88
99/100
87/99
75/100
0.25
96/101
81/99
84/99
0.07
121/100
70/101
66/97
0.36
58/101
91/105
108/92
0.34
93/112
66/91
102/96
0.80
1.00
1.30
1.32
1.00
0.89
0.77
1.00
0.86
0.88
1.00
0.57
0.55
1.00
1.49
2.02
1.00
0.88
1.28
0.85,1.98
0.86,2.03
0.59, 1.33)
0.51,1.17
0.57,1.29
0.58, 1.32
0.38, 0.85
0.36, 0.83
0.97,2.30
1.31,3.12
0.58, 1.35
0.86, 1.91
'Adjusted for matching covariates, and body mass index(linear)
66
Table 2.9. Associations of CRP and cytokines with Benign Prostatic Hyperplasia risk,
stratified by age'
CRP, (,mg/L)
Tl
T2
T3
TNF-a, (pg/mL)
Tl
T2
T3
No. cases/
controls
120/131
118/132
151/130
137/133
142/132
114/130
sTNF-RI, (pg/mL)
Tl
T2
T3
119/133
123/130
151/132
sTNF-RII, (pg/mL)
Tl
T2
T3
EL-6, (pg/mL)
Tl
T2
T3
IFN-Y, (pg/mL)
Tl
T2
T3
202/132
92/132
95/129
91/134
102/131
196/128
109/118
124/149
160/128
Age 55-65
OR
1.00
0.96
1.23
1.00
1.05
0.84
1.00
1.03
1.26
1.00
0.45
0.48
1.00
1.15
2.24
1.00
0.88
1.34
(95% CI)
(0.68, 1.37)
(0.86,1.80)
P =
-* interaction
(0.75, 1.18)
(0.59, 1.18)
p. . =
-* interaction
(0.72,1.46)
(0.90,1.78)
P • • =
J
interaction
(0.32, 0.64)
(0.34, 0.68)
P • =
1
interaction
(0.80, 1.67)
(1.57,3.18)
P • • =
J
interaction
(0.61, 1.25)
(0.94, 1.90)
P =
-* interaction
No. cases/
controls
75/96
91/93
110/95
0.53
93/97
94/96
96/95
0.30
103/98
78/95
102/95
0.41
98/97
81/95
100/95
0.02
82/97
100/102
94/88
0.03
105/114
73/81
105/93
0.71
Age 65+
OR
1.00
1.18
1.36
1.00
1.05
1.12
1.00
0.78
0.98
1.00
0.85
1.03
1.00
1.14
1.17
1.00
1.03
1.25
(95% CI)
(0.77, 1.81)
(0.89, 2.07)
(0.70, 1.59)
(0.74, 1.68)
(0.51, 1.17)
(0.66, 1.47)
(0.56, 1.28)
(0.69, 1.54)
(0.76,1.70)
(0.77, 1.79)
(0.68, 1.57)
(0.85, 1.86)
a
Adjusted for matching covariates, and body mass index(linear)
67
Table 2.10. Associations of CRP and cytokines with Benign Prostatic Hyperplasia
risk, stratified by type of BPH defining event
3
CRP, (mg/L)
Tl
T2
T3
TNF-a, (pg/mL)
Tl
T2
T3
Type of BPH
BPH Treatment
b
No. cases/
controls
111/227
109/229
155/227
101/231
102/225
98/227
sTNF-RI, (pg/mL)
Tl
T2
T3
95/230
89/225
117/228
sTNF-RII, (pg/mL)
Tl
T2
T3
BL-6, (pg/mL)
Tl
T2
T3
IFN-y, (pg/mL)
Tl
T2
T3
175/232
91/223
109/228
99/243
109/207
167/233
89/190
102/273
110/220
OR
1.00
0.93
1.08
1.00
0.98
0.90
1.00
0.86
1.16
1.00
0.53
0.67
1.00
1.52
1.60
1.00
1.04
1.33
(95% CI)
(0.66, 1.32)
(0.77, 1.53)
p
c

-* difference
(0.72, 1.34)
(0.66, 1.23)
p
c

r
difference
(0.62, 1.19)
(0.85, 1.58)
p
c
=
•* difference
(0.38, 0.75)
(0.48, 0.92)
p
c

•» difference
(1.07,2.16)
(1.13,2.25)
P
c

r
difference
(0.75,1.43)
(0.96,1.85)
p
c

r
difference
Defining Event
BPH
No. cases/
controls
97/227
92/229
112/227
0.27
133/231
127/225
115/227
0.63
123/230
107/225
145/228
0.74
140/232
71/223
90/228
0.71
77/243
100/207
124/233
0.67
93/190
140/273
142/220
0.27
Sympi
OR
1.00
0.94
1.29
1.00
1.04
0.99
1.00
0.97
1.24
1.00
0.54
0.62
1.00
1.27
1.68
1.00
0.81
1.06
toms
b
(95% CI)
(0.68, 1.31)
(0.94, 1.78)
(0.74, 1.45)
(0.71, 1.38)
(0.68, 1.36)
(0.89, 1.73)
(0.39, 0.74)
(0.46, 0.84)
(0.91, 1.77)
(1.22,2.31)
(0.57, 1.14)
(0.75, 1.49)
"Adjusted for matching covariates, and body mass index(linear)
'Treatment includes men who received drug or surgical intervention(n=322); BPH Symptoms includes men with
two IPSS scores >14 or two IPSS scores at least 5 units higher than baseline plus at least one score >12 (n=405)
c
Pdiffetence is calculated as the difference in risk between BPH Treatment and BPH Symptoms as estimated from a
polychotomous logistic model
68
Table 2.11. Associations of CRP and cytokines with Benign Prostatic Hyperplasia
risk, stratified by time from baseline to BPH event'
Time from baseline to BPH event
CRP, (mg/L)
Tl
T2
T3
TNF-a, (pg/mL)
Tl
T2
T3
No. cases/
controls
34/227
38/229
48/227
37/231
41/225
42/227
sTNF-RI, (pg/mL)
Tl
T2
T3
33/230
43/225
44/228
sTNF-RII, (pg/mL)
Tl
T2
T3
IL-6, (pg/mL)
Tl
T2
T3
IFN-y, (pg/mL)
Tl
T2
T3
57/232
26/223
37/228
27/243
42/207
51/233
30/190
47/273
43/220
0 - 3 years
OR
1.00
1.11
1.30
1.00
1.20
1.19
1.00
1.35
1.30
1.00
0.47
0.68
1.00
1.86
1.89
1.00
1.11
1.26
(95% CI)
(0.67, 1.84)
(0.78,2.14)
p *-
r
difference
(0.74,1.95)
(0.73, 1.94)
p
b
-
r
difference
(0.82, 2.22)
(0.79,2.15)
p
b
-
* difference
(0.29, 0.79)
(0.43, 1.07)
p
b
-
r
difference
(1.10,2.16)
(1.12,3.17)
p
b
-
r
difference
(0.67, 1.83)
(0.75,2.10)
p
b
-
*• difference
No. cases/
controls
174/227
163/229
219/227
0.69
197/231
188/225
171/227
0.25
185/230
153/225
218/228
0.78
258/232
136/223
162/228
0.84
149/243
167/207
240/233
0.68
152/190
195/273
209/220
0.89
4 - 7 years
OR
1.00
0.91
1.18
1.00
0.97
0.89
1.00
0.83
1.17
1.00
0.55
0.63
1.00
1.29
1.60
1.00
0.89
1.18
(95% CI)
(0.68, 1.21)
(0.89, 1.56)
(0.74,1.28)
(0.68,1.18)
(0.62,1.10)
(0.89, 1.54)
(0.41,0.73)
(0.48, 0.83)
(0.97, 1.73)
(1.21,2.12)
(0.67, 1.18)
(0.89, 1.58)
"Adjusted for matching covariates, and body mass index(linear)
b
p-difference is calculated as the difference in risk between BPH Treatment and BPH Symptoms as estimated from
a polychotomous logistic model
69
1
I
I
1
3
--I
P ©
. 2 6*
-2
c
03 O
•3 S
a < H
CN P
< N ^
U O
^ &
EA O
° s
13 co -
C3
o
6
? °
l « -to*
£
s
O V
§ 1
2 'C
.S >i
Si
® «
U
£ S
S3 ©
•c &
O
° a
I
8
TO u,
°S
| 8
—t OO
r o O
rt" CN
o ,—T vo" °°
O If i CO ^
^ O O O —
—< Tf
oo m
CN
CN
CN CN
CN CN
O
O
T f OO
os" vo" ^
vn t~- J*.
o o °
CN OO
CN
CN
ON
CN CN
CN CN
0 0 ^ H
TT VO
o
o
Ov O
m VO
„ - oo
£ s 2
o o °
—> VO
O -H
r-
CN
CN
CN CN
CN CN
VO VO
—i < *
T3
a
oo >/->
o o ^
_ - ~ ©
© v~> m o
P m cs (q
2- 2- °
rn vo
> /-> rn
o o
N n oo
e n CN CN
CN CN CN
f^ 3\ f^
t-~ m CN
VO rt
oo o
o ~
o o
«/-> r -
«/-> vo
o
o
-
1
>/ ->
m
CN
0 \
o ,
oo
CN
© ,
oo
•«a-
o
o
o
o
VO
O
O
oo CN
oo o
~ CN
t ^ ""> CN
C- OO "_•
o o °
O O -H
O CN ro
AH J
CN m oo
m CN CN
CN CN CN
CN OO OO
c- ro " *
Tf CN
t-~ O
o ^
ON VO o
n i n •
© o °
^—' ^—'
o • * vo
o •/-> r-
c i CN c j
i f i J ; ? i o
vo ^ CN =S
o
CN CN CN
vo oo -rr
TJ- CN vo
VO
J N rt
- ^" °
o ^- o
o n p
^ V
O T f
<3\
0 0
CN
o
ON
O
CN CN •!-•
"3D H
CN m
a"
oX
^ f _ H H
>—i CN m
H H H
3
n « «
w . y
J i - r - t O
70
0>
S 55
I—I
1 0
O N
O
V5
NO
i
« r>
8?
s
_ i
* ^ --.
o 8. 3
5 8
u w 5
s o -^
°& *
l l
J "E "g J3
° « 6 °
=
PQ
•a S
es o
j5»ce -
u
ON
o
u . 3
M P
o
CN ON
V~I OO
oo o
Tj - NO
o >/-> r-~
o oo o
rt O rt
CN <N —<
co co co
CN N O —i
n N w
oo u-i
o p. °
O N Nt
O O Tf
t N CN ~
co co co
p - od oo
CN CN co
0 0 ON
>/-N OO
S § ">
o o. °
O co co
O O tN
CN CN - H
CO C I CO
T3
• -I *T> T f
NO NO OO
NO CN
0 0 NO
d d ^r
~ o
O OO o
CO —• o
2, 2 ©
o o •<*•
O VN. CO
—! d d
NO ON O
CN CO CO
NO NO
CN ^ H
CN" oo
NO NO
d o
t
d" CN" §
CN CN o
2- 2- °'
O !/ "> 0O
O CO CO
^' d d
NO ON O
CN CO CO
l/-> CN CN
NO OO
& d ~
- CO
oo h g
CN CO P
2 2 °
O CN NO
O T f K0
-H d d
NO
CN
ON O
CO CO
a
0)
is
CN
o
a, j
o
NO
5* ^* CN co
M H H H
3 -* CN CO
r-~ CN
CN • *
^ H " O "
r- CN
O ON NO
O CO CN
oo ON OO
o co ^r
t < NO
CN Tj -
oo >/->
—< CN
- < -" o
CO OO ~ -
2 2
O —i CN
O NO W-N
-H" d ^
oo ON oo
O CO • *
^ t OO r H
CN <-" </ •>
CN •<*•
ON uo
^ " N
CN" CN" §
t - "fr o
d ^
O OO V)
O - H CN
OO ON 2 g
o co 2 is
CO l/-> ^
S3
M "
H
OX)
NO"
I
^ CN CO
H H H
CZ!
OH
"c3
o
• * - >
o
a
tfl « s o . s
o o g- c
B c e c
.2 .2 « .£
t: -e o t
o o > c
a, Q.'S c
O O tS C
PL, P -i ^j P .
71
I
O
s .
S .
s
o
©
•a
4>
•a
7!
e
Si
w s
"S C
OS 0
•M •
I »
o
l*
CM
s
0J5
S
ai
PQ
•O
s
«
o
s
v
A
^ O
41
o
3
fa
</-s
as
s© Qj
fa
Z
H
</-s
o
£
Pi
O
U
Os
p i
o
u
fa ^
&
-a
o
PQ
O
U
Os
O
U
OS
04
O
B
© o\
Os oo
o o
© OO CN
O —i CN
OO
r- ;_
8
o
VO
•*• o
OS Os
©
o
CN CN
a
©
fa
Z
H
CO O)
CD W
o
o CN TT
s
©
SO OO
OS OS
O Tf O
© CN m
©
o
CI
SO

t ~
Os
o
so
CN
C~
oq
1—1

o
p
, _ 4
c~
r o
«"s
o
o
II
T3
B
o
r

t
r
e

<*H
D,
u
OO —I
d oo
Os OO
CN CN
OO CN 2
N OO 1
^4 ^4 o
- - II
os os a
O O £
o
©
fa
H
W5
c
es
« /-s
> /-s
*-> —i a
OS Os p
p © •"
Js O O •-< OS CN so O,
a!
u
cs
es
H
J3
60
'5
£
13
o
£
60
'5
j r
>
O
.o
o
M
'5
PS
"c3
o
£
5
60
'S
p
>
o
^)
o
JS
60
'S
^

o
^
£
60
O
v
e
r
w
e
i
,

e
s
e

43
O

of
rt
f
o
r

u
s
t
e
d

^
<
k i
60
A3
f
o
r

u
s
t
e
d

^
<
72
Chapter 3
Prostatic Hyperplasia and Risk of Prostate Cancer
Abstract
This study examined the association between benign prostatic hyperplasia
(BPH) and prostate cancer risk in 5,068 placebo-arm participants of the Prostate
Cancer Prevention Trial (1993-2003). These data include 1,225 men who had cancer
detected during the 7-year trial, 556 which were detected following abnormal PSA or
DRE ('For-cause') and 669 detected without indication ('Not-for-cause'), and 3,843
men who had biopsy proven absence of prostate cancer at the trial end. BPH was
assessed hierarchically as self-report of surgical or medical treatment, moderately
severe symptoms (average recruitment and baseline International Prostate Symptom
Score (IPSS)> 14), or physician diagnosis only at baseline, and analyses were
completed using BPH status at baseline ('prevalent') or any evidence of BPH prior to
cancer diagnosis or end of study biopsy ('prevalent plus incident'). Controlled for age,
race and body mass index, there was no association of 'prevalent' nor 'prevalent plus
incident' symptomatic BPH with prostate cancer risk. This lack of association was
consistent across subgroups defined by type of BPH defining event (treatment,
symptoms or physician diagnosis), prompt for prostate cancer diagnosis and prostate
grade. This study provides the strongest evidence to date that BPH does not increase
the risk of prostate cancer.
Introduction
Benign prostatic hyperplasia (BPH) and prostate cancer are common urological
73
conditions in older men and there is a large body of evidence supporting a relationship
between these conditions. Original hypotheses of an association between BPH and
prostate cancer were based on studies documenting the presence of both conditions at
autopsy [82-84]. In more recent studies, BPH and prostate cancer frequently coexist;
more than 20% of men with prostate cancer also have BPH and cancer is found
incidentally in a significant percentage (10-20%) of surgically removed BPH
specimens.[88,89,101,123] In addition, epidemiologic studies have reported an
association between BPH and prostate cancer[95,96] There are also several similarities
between the two conditions: the prevalence of both BPH and prostate cancer increase
in parallel with age[88]; both conditions require androgens for growth and
development[88,90]; and both respond to androgen-deprivation treatments.[91,92]
Gene expression studies have identified similarities in genetic alterations between BPH
and prostate cancer, particularly in the growth regulatory genes[93]. An expanding
body of evidence also supports an important role and inflammation in both. [78,94]
Despite these similarities, there are also strong arguments that BPH and prostate cancer
are unrelated, both clinically and biologically. Differences in the histology and
anatomical location of these two conditions fuel the current opinion that BPH and
prostate cancer are not linked in origin.[87] BPH is characterized by hyperplasia of
primarily stromal and, to a lesser extent, epithelial cells, whereas prostate cancer
involves hyperplasia of only the epithelium in the glandular compartment of the
prostate[5,86]. Nearly all BPH arises in the transition zone of the prostate, while only
74
24% of prostate cancers occur in the transition zone, two-thirds in the peripheral zone,
and the rest in the central zone.[85]
From the perspective of epidemiologic research, the most substantial problem
in interpreting studies of the association between BPH and prostate cancer is the
potential for detection bias. Autopsy studies have shown that clinically significant
prostate cancer is undiagnosed in up to 15.6 percent of men[124]; therefore any
condition that increases the rate of urologic evaluation or PSA testing is likely to
increase the detection rate of prostate cancer that would otherwise have gone
undetected. Symptoms of BPH can increase the detection rate of prostate cancer by
increasing the number and extent of urologic examinations^ 100,125] and treatment of
BPH via trans-uretheral resection of the prostate (TURP) can result in the detection of
incidental prostate cancer. [101] In addition, BPH can increase prostate specific antigen
(PSA) levels, thereby increasing the likelihood of detecting latent prostate cancer. [102]
BPH also causes an enlargement of the prostate gland which could lead to increased
biopsies due to abnormal digital rectal examination[126], although increased prostate
size could also make it more difficult to detect cancer[103].
The Prostate Cancer Prevention Trial (PCPT) offers a unique opportunity to
investigate the association between BPH and prostate cancer. In the PCPT, all men
were subject to standardized diagnostic scrutiny for prostate disease regardless of
symptomatology; every participant completed an annual assessment of lower urinary
tract symptoms and received annual DRE and PSA tests. Furthermore, all men who
75
were not diagnosed with prostate cancer during the trial were asked to undergo an end-
of-study biopsy to confirm the absence or presence of prostate cancer. Here we give
results of a prospective study examining the association between BPH and prostate
cancer risk among placebo-arm participants in the PCPT. This study addresses many of
the limitations of previous studies, and allows a rigorous test of the association
between BPH and prostate cancer risk.
Materials and Methods
Data are from the Prostate Cancer Prevention Trial (PCPT), a randomized,
placebo-controlled trial testing whether finasteride reduced prostate cancer risk. [92]
Briefly, 18,880 men age 55 years and older with normal digital rectal exam (DRE),
prostate-specific antigen (PSA) levels of <3 ng/ml, no history of prostate cancer or
other clinically significant coexisting conditions, and no severe BPH symptoms,
defined as an International Prostate Symptom Score (IPSS)[6] of 20 or higher, were
randomized to receive finasteride (5 mg/day) or placebo.
During the PCPT, men underwent DRE and PSA determinations annually. A
prostate biopsy was recommended for participants who had an abnormal DRE or a
PSA >4.0 ng/mL. At the final study visit in Year 7, all men who had not previously
been diagnosed with prostate cancer were offered an end-of-study (EOS) biopsy. All
biopsies were collected under transrectal ultrasonographic guidance and involved a
minimum of 6 specimens (cores). All biopsies were reviewed to confirm diagnoses of
adenocarcinoma by both the pathologist at the local study site and by a central
76
pathology laboratory. Tumors were graded centrally using the Gleason scoring system.
Cancers were categorized as low-grade (Gleason sum, 2 to 7 (3+4 only)), and high-
grade (Gleason sum 7 (4+3 only) to 10) to distinguish those more likely to be clinically
significant. Prostate cancers were also categorized by prompt for diagnosis: cancers
detected following abnormal PSA or DRE were classified as 'For-cause'; and cancers
detected without indication, primarily at the study exit biopsy, were classified as 'Not-
for-cause'.
Study Population
This study is conducted in the 9,457 placebo arm participants because
finasteride prevents both BPH and prostate cancer[92]. We first excluded 3,637 men
who did not have a prostate cancer diagnosis or end-of-study biopsy, which included
585 men who died, 2,804 men who were medically unable or refused and 248 men
who, due to early completion of the trial (June 24, 2003) were not yet eligible for the
end-of-study biopsy. We then excluded 183 men whose end-of-study biopsy was
completed > 180 days before or after their planned end-of-study visit, and 376 men
whose end-of-study biopsy was completed after the end of the trial. Prostate cancer
events that occurred > 180 days after the planned end-of-study visit or after the end of
the trial were ignored. From the 5,261 remaining eligible men (1,264 cases; 3,997 non-
cases), we excluded 85 men (19 cases, 66 controls) who reported taking anabolic
hormones (including testosterone gel/injection/patch, nandrolone or
dehydroepiandrosterone) and 40 (9 cases, 31 controls) men who reported taking
77
finasteride during the PCPT. Furthermore, we excluded 68 (11 cases, 57 controls) men
who reported treatment (3 surgery, 65 medical) for BPH during the PCPT without
evidence of lower urinary tract symptoms (LUTS) (IPSS>10) or physician diagnosis of
BPH, whose surgery or use of alpha blockers was likely unrelated to BPH, leaving
5,068 men (1,225 cases, and 3,843 controls) for these analyses.
Data Collection
At baseline, age, race/ethnicity, family history of prostate cancer in first-degree
relatives, physical activity, usual alcohol consumption, and history of smoking were
collected using self-administered questionnaires and height and weight were measured
by clinic staff. Body mass index (BMI) was calculated as weight (kg) divided by
height
2
(m). Extensive medical data, including physician diagnosis of and treatment
for BPH was collected at the baseline, 6-month, and annual clinic visits and at every 3-
and 9-month phone contact between clinic visits. At recruitment, baseline, and each
annual clinic visit, participants completed the IPSS[6], a 7-item self-administered
questionnaire assessing the frequency of LUTS. DRE and PSA were assessed at
baseline and each annual clinic visit, and prostate volume was assessed by trans-rectal
ultrasound at the time of diagnostic biopsy (for interim cancers) or the end of study
biopsy.
Prevalent BPH (at baseline) was defined hierarchically as self report of surgical
(TURP, balloon dilation or prostatectomy) or medical (tamsulosin, doxazosin,
terazosin) treatment for BPH, presence of moderately severe LUTS (average IPSS from
78
recruitment and baseline > 14) or self-report of physician diagnosis of BPH at baseline.
Incident BPH (post-baseline) was defined hierarchically as the most severe BPH
defining event (surgical or medical treatment, moderately severe lower urinary tract
symptoms (2 EPSS > 14) or self-report of physician diagnosis of BPH) during follow-
up prior to cancer diagnosis or end-of-study biopsy.
Methods
Although data are from a clinical trial, prospective data analyses are
inappropriate because the presence/absence of prostate cancer was not ascertained
annually, and for most men, was determined at the end-of-study biopsy. Therefore, the
seven-year period prevalence of prostate cancer was used as the primary outcome.
Relative risk regression methods[127] were used to evaluate the association
between BPH and prostate cancer, since biopsy-confirmed prostate cancer was
common (24.4% of men) and thus the odds ratio would not be a good estimate of the
relative risk. [92] Risk ratios were estimated using the SAS GENMOD procedure with
Log-link function and binomial distribution[128]. When log-binomial models did not
converge, the GENMOD procedure was used with a Poisson distribution, and robust
error variances were estimated using a repeated statement and the subject identifier as
the class variable[128].
Multivariate models were used to estimate associations of BPH with risks of
total, and low-grade and high-grade prostate cancer while controlling for covariates.
Analyses were completed using BPH status at baseline ('prevalent' BPH) or any
79
evidence of BPH preceding cancer diagnosis ('prevalent plus incident' BPH) as the
exposure. Associations of incident BPH and prostate cancer were not investigated
alone due to recency of exposure and smaller sample size. Results are given adjusted
for age (linear), and race (white, black, and other) (Model 1), and body mass index
(continuous) (Model 2). Additional control variables associated with BPH or prostate
cancer risk in this sample, including alcohol consumption[107,129] and smoking status
[130], did not affect risk estimates and were not included in the final models. Models
were not controlled for prostate volume because it was missing differentially among
men diagnosed with interim cancer.
For both 'prevalent' and 'prevalent plus incident' BPH we investigated the
individual associations between type of BPH defining event (treatment, symptoms or
physician diagnosis) and prostate cancer risk. To assess potential biases, additional
analyses investigated the associations of BPH with 'for-cause' and 'not-for-cause'
prostate cancer. All statistical tests were 2-sided and were considered statistically
significant at P<0.05. Statistical analyses were conducted using SAS software (version
9.1; SAS Institute, Inc., Cary, NC).
Results
Among the total 5,068 men in this analysis, 1,549 (30.6%) had BPH at baseline:
286 (18.5%) had surgical (281) or medical (5) treatment for BPH; 276 (17.8%) had an
average of 2 IPSS>14; and 987 (63.7%) had a physician diagnosis only. Post-
randomization, 763 men had an incident BPH defining event: 350 (45.9%) had surgical
80
(31) or medical (319) treatment; 329 (43.1%) had 2 IPSS>14; and 84 (11.0%) reported
a physician diagnosis of BPH. The median age of participants was 64.9 ± 5.4 years,
and most men were Caucasian, former or never smokers, overweight, light or
moderately active, and were not heavy drinkers (Table 3.1). Most participants also had
a baseline PSA <1.1 ng/ml, did not have a family history of prostate cancer, and had a
prostate volume <3 3.5 cm (Table 3.1).
'Prevalent' BPH was more common among older men, men of 'other' race,
men who did not smoke, were not overweight or obese and were not heavy drinkers,
men with a higher baseline PSA and IPSS, and a larger prostate volume (Table 3.2).
'Prevalent plus incident' BPH was more common among older men, men who were not
heavy drinkers, men with a higher baseline PSA and IPSS, and a larger prostate
volume (Table 3.3).
Among the total 5,068 men in this analysis, 1,225 (24.2%) were diagnosed with
prostate cancer: 1,074 (87.7%) had Gleason scores from 2 to 7 (3+4), 105 (8.6%) had
Gleason scores >7 (4+3) and 46 (3.8%) had unknown Gleason scores. 556 (45.4%)
cancers were detected following abnormal PSA or DRE and were classified as 'For-
Cause', while 669 (54.6%) were diagnosed at the end-of-study biopsy without clinical
indication ('Not-For-Cause'). Prostate cancer was more common among older men,
African-American men, men who were not overweight or obese, men with a higher
baseline PSA, and men who had a family history of prostate cancer (Table 3.4).
Tables 3.5 and 3.6 give associations of 'prevalent' and 'prevalent plus incident'
81
BPH with cancer risk. There were no associations of prevalent BPH or prevalent plus
incident BPH with prostate cancer risk. Results were similar when BPH was
categorized by type of BPH defining event (surgery, symptoms, physician diagnosis),
and when prostate cancer was categorized by grade (low, high) and prompt for
diagnosis ('For-cause', 'Not-for-cause'). Associations were similar when controlled
for age and race only (Model 1) and controlled in addition for body mass index (Model
2). There were no substantial differences in results when analyses excluded 47 prostate
cancer cases occurring within 1 year of the first BPH defining event, or when high
grade cancers were restricted to Gleason sum 8+ (3.7, 3.8).
Discussion
In this large prospective study among placebo-arm participants in the PCPT
there were no associations of symptomatic BPH with prostate cancer risk. This lack of
association was consistent across subgroups defined by type of BPH diagnosis
(treatment, symptoms or physician diagnosis), prompt or reason of prostate cancer
diagnosis ('For-cause' or 'Not-for-cause') and prostate cancer grade (Gleason score 2-
7(3+4) vs. 7(4+3)-10).
Six previous studies examined associations between BPH and prostate cancer
risk. These include: two hospital-based case-control studies that estimated the risk of
prostate cancer among men with BPH compared to men hospitalized for non-
neoplastic diseases[95,97]; 2 studies that calculated standardized incidence and/or
mortality ratios by comparing the incidence/mortality of prostate cancer in a cohort of
82
men hospitalized with BPH to the general population[96,98], and two studies that
compared prostate cancer incidence or mortality in a cohort of men hospitalized with
BPH to a cohort of men hospitalized with non-neoplastic diseases[95,99]. Of these, 3
studies reported an increased risk of prostate cancer. The first was a case-control study
that reported an increased risk of prostate cancer among men previously hospitalized
for non-cancer prostatic disease[95], the second was a prospective study that reported
an increased risk of prostate cancer mortality among men hospitalized with a primary
diagnosis of BPH [95], and the third was a prospective study that reported an increased
prostate cancer incidence among men who had a TURP as well as increased prostate
cancer incidence and mortality among men with a diagnosis of BPH without surgical
treatment [96].
Bias is a likely explanation of all the previously published studies showing an
association between BPH and prostate cancer. The largest single source of bias in
these studies is due to the increased likelihood of prostate cancer detection in men with
symptomatic urinary disease. Men with more severe urinary symptoms and larger
prostate size seek more health care for urinary symptoms[100,131]. In studies
investigating the association between BPH surgery and prostate cancer risk, men being
considered for surgical treatment of BPH are more likely to be thoroughly evaluated
for prostate cancer because surgical treatment of BPH can increase the chance of
detecting latent prostate cancers[101]. This may explain the findings in studies by
Chokkalingam and Armenian, in which the initial significant elevations in prostate
83
cancer incidence[96] and mortality [95]in the first few years of follow-up after initial
BPH diagnosis decrease with increasing duration of follow-up.
Additional issues also contribute to bias in previously published studies,
including differential exclusion criteria between men with and without BPH, and use of
the general population as a comparison group. In the prospective study by Armenian et
al., the exclusion of men at high prostate cancer risk (positive DRE) from men without
BPH likely resulted in a spuriously low rate of prostate cancer among men without
BPH, and the inclusion of men at high prostate cancer risk among men with BPH
(positive DRE) resulted in a spuriously high rate of prostate cancer among men with
BPH; thus yielding an increased prostate cancer mortality[95]. In addition, the
differential exclusion of men with latent prostate cancer from men with surgically
treated BPH without a similar exclusion for either men with BPH symptoms only or
from the contrast group of men without BPH introduces bias[96]. Furthermore, the use
of the general population as a comparison group without control for established
prostate cancer risk factors (age, race, obesity and physical activity), is likely to
introduce bias since the distribution of these factors among men with BPH differs from
that in the general population. [96] Overall, given both detection bias and shortcomings
in study design, we do not feel that previous studies of BPH and prostate cancer risk
could validly test whether or not there was a true association.
To evaluate the extent to which detection bias could have affected the results of
our study, we completed analyses stratified by whether prostate cancer was diagnosed
84
due to an elevated PSA or abnormal DRE ('For-cause') or was diagnosed without
clinical indication ('Not-for-cause'). If the presence of BPH was increasing the
likelihood of prostate cancer detection, we would expect BPH to be associated with an
increased risk for 'For-cause' prostate cancers. There was a suggestion of increased
risk between prevalent BPH and 'For-cause' prostate cancers diagnosed within the first
four years of the PCPT (OR=1.16, 95% CI: 0.82, 1.66), compared to cancers diagnosed
in years 5-7 (OR=1.03; 95%CI: 0.84, 1.25), which is consistent with detection bias;
however, neither association was statistically significant.
The evaluation of BPH and prostate cancer risk without control for PSA or
prostate volume deserves comment. PSA is produced by the prostate proportionally to
prostate volume, and the benign enlargement characteristic of BPH is associated with
increased PSA levels. [102] However, surgical treatment of BPH, which removes
tissue from the transition zone of the prostate, also lowers PSA levels[132]. In our
study population, baseline PSA was strongly associated with both total BPH and
prostate cancer (Tables 1 and 2), though for men with BPH treatment at baseline, 98%
of which were BPH surgery, the distribution of PSA at baseline was much lower than
for men with symptoms or physician diagnosis (Table 1). We assessed a model
adjusted for baseline PSA in addition to the covariates in model 2; which resulted in a
risk ratio of 0.96 (95% CI 0.85, 1.09) for the association between total prevalent BPH
and total cancer. However, this model yielded a spurious positive association between
'prevalent' BPH surgery and prostate cancer risk, while minimizing the associations
85
between BPH symptoms and physician diagnosis and prostate cancer risk. This model
is inappropriate and leads to biased results because adjustment for baseline PSA has a
differential affect on the associations between type of BPH defining event and prostate
cancer risk. We also considered controlling models for prostate size, because an
enlarged prostate could lead to increased biopsies due to abnormal digital rectal
examination[126] or could also make it more difficult to detect cancer. [103] However,
prostate size tended to be measured more frequently at end-of-study biopsies and less
frequently during interim biopsies, and because high grade disease was more likely to
be detected on interim biopsies, prostate volume was also missing more frequently for
high grade cancers (prostate volume was missing for: 42% of cancers detected during
interim biopsies, 10% of cancers detected at the end-of-study biopsy, and 31% of high-
grade, 21% of low-grade cancers). A model adjusted for prostate volume would be
problematic because high-grade prostate cancers are differentially missing.
There are several elements of the PCPT study design which serve to minimize
bias in this study of BPH and prostate cancer risk. First, men with possible prostate
cancer, defined as abnormal DRE or PSA>4, were excluded from participation in the
PCPT. Second, during the PCPT all men were subject to the same diagnostic scrutiny
for prostate disease (both BPH and prostate cancer), regardless of symptomatology;
every participant received an annual DRE and PSA test, and completed an assessment
of lower urinary tract symptoms, and men with an abnormal DRE or PSA were
recommended for prostate biopsy. Therefore, by standardizing screening of all men for
86
prostate disease, the PCPT minimized the impact of BPH symptoms on the number and
extent of urologic exams. Third, all men who were not diagnosed with prostate cancer
during the trial were asked to undergo an end-of-srudy biopsy to confirm the absence
or presence of prostate cancer. Although men with BPH were more likely to have a
prompt for biopsy due to abnormal DRE and/or PSA, the end-of-study biopsy provided
an equivalent opportunity to detect latent prostate cancer in men who had no biopsy
prompt. Therefore, the restriction of our study population to men with a known cancer
status minimized detection bias. The restriction of our study population to men with
known cancer status also reduced the possibility that undetected cancers could affect
study results. Lastly, the PCPT collected data that allowed us to investigate the
association between BPH and prostate cancer risk separately for subgroups defined by
prompt for BPH diagnosis, including physician diagnosis, symptoms and surgical and
medical treatment, and this is the only study to incorporate medical treatment in
definitions of BPH. In addition, all cancers were graded centrally, and this is the first
study to look at cancers stratified by grade.
There are several limitations to this study. Our analyses were restricted to men
with known cancer status; therefore, men who did not complete the end of study
biopsy, for whom cancer status was unknown, may differ from this study population.
For example, men who had a negative interim biopsy were more likely to have an end
of study biopsy (73.0% vs. 62.4% of men who did not have a negative interim biopsy;
p<0.0001), and therefore were more likely to be included in these analyses. However,
87
the completion the end-of-study biopsy was not related to BPH status (62.4% of men
with and 64.4% of men without BPH completed the end-of-study biopsy); therefore
this study population is likely representative of the overall PCPT population. In
addition, the relatively strict eligibility criteria of the PCPT (no major co-morbidities,
PSA less than 4 and IPSS less than 20) yielded a unique study population which limits
the generalizability of findings.
Conclusion
This study confirms the current opinion that BPH is not associated with prostate
cancer risk. This is the only study to report on the association between BPH and
prostate cancer risk in a study population where diagnostic surveillance of both BPH
and prostate cancer was rigorous and complete, and thus provides the strongest
evidence to date that BPH does not increase the risk of prostate cancer.
88
Table 3.1. Distribution of demographic and lifestyle
characteristics among the total study population
Total Sample
N=5,068
n %
Age
55-59
60-64
65-69
70+
Race
Caucasian
African American
Other
Smoking Status
Current
Former/Never
Body mass index(kg/m
2
)
Normal (<25)
Overweight (25-29.9)
Obese (>30)
Physical activity
Sedentary
Light activity
Moderate activity
Very active
Alcohol (grams/day)
a
<0.25
0.25- 1.74
1.75- 12.9
13-26.9
>27
Baseline PSA (ng/ml)
b
0-1.0
1.1-2.0
2.1-4.0
Family history prostate
cancer
c
Yes
No
Baseline IPSS
d
0-3.5
3.6-6.5
6.6-14.0
14.1-20.0
1291 32.0
1229 31.8
849 22.9
474 13.3
3603 93.6
101 3.0
139 3.4
268 6.9
3571 93.1
995 26.9
1978 51.8
826 21.3
633 16.5
1609 41.7
1198 32.0
385 9.8
998 26.1
642 16.9
1275 32.6
551 14.3
376 10.1
2002 46.9
1292 35.5
549 17.6
609 17.1
3234 82.9
1052 27.3
990 25.3
923 24.1
878 23.4
89
Table 3.1. Continued. Distribution of demographic and
lifestyle characteristics among the total study population
Total Sample
N=5,068
n %
Prostate volume (cm
3
)
e
10.0-18.8 cm
3
18.9-25.1 cm
3
25.2-33.5 cm
3
33.6-100.0 cm
3
846
945
834
794
25.2
28.0
24.2
19.6
* <0.25 grams/day(<l drink/month), 0.25-1.74 grams/day(l-3 drinks/month),
1.75-12.9 grams/day(l-6 drinks/week), 13.0-26.9 grams/day(7-13
drinks/week), >27 grams/day( >14 drinks/week)
b
Men with PSA>4.0 at baseline were excluded from participating in the
PCPT
c
Self-report of father, brother or son having prostate cancer
d
Average IPSS from recruitment and baseline visits, men with IPSS at
baseline > 20 were excluded from participating in the PCPT
e
Excludes 626 men missing prostate volume and 76 men with prostate
volume <10 or>100; prostate volume captured at the EOS prostate biopsy
(men without cancer) or diagnostic biopsy (men with cancer)
90
Table 3.2. Percent of men with prevalent BPH by demographic and lifestyle characteristics
Age
55-59
60-64
65-69
70+
Race
Caucasian
African American
Other
Smoking Status
Current
Former/Never
Body mass index(kg/m )
Normal (<25)
Overweight (25-29.9)
Obese (>30)
Physical activity
Sedentary
Light activity
Moderate activity
Very active
Alcohol (grams/day)
f
<0.25
0.25- 1.74
1.75-12.9
13-26.9
>27
Baseline PSA (ng/ml)
8
0-1.0
1.1-2.0
2.1-4.0
Family history prostate cancer
11
Yes
No
Total BPH
n=
%
22.6
28.1
37.5
43.6
30.8
19.5
33.5
21.9
31.2
33.7
30.4
26.7
31.8
30.6
28.9
34.0
34.8
31.5
28.9
28.6
26.3
26.0
32.6
38.5
30.6
30.2
=1,549
P
b
O.0001
0.01
0.0003
0.001
0.14
0.0006
O.0001
0.78
Prevalent BPH"
Type
Treatment'
n=286
%
1.9
3.9
8.1
14.7
5.6
3.3
9.4
2.6
5.9
6.9
5.2
5.0
5.2
5.9
5.8
4.9
7.1
5.5
5.8
4.0
4.1
6.6
4.9
4.5
5.7
5.6
of BPH defining
Symptoms'
1
n=276
%
5.0
5.2
5.8
6.5
5.5
2.0
8.2
5.1
5.5
5.3
5.8
4.9
5.9
5.5
4.9
6.3
6.3
5.5
5.8
4.0
4.1
4.7
5.5
7.4
4.7
5.6
event
Physician
Diagnosis
6
n=987
%
15.7
19.0
23.7
22.4
19.8
14.3
15.9
14.3
19.8
21.5
19.3
16.8
20.7
19.2
18.2
22.9
21.5
20.7
17.4
19.4
19.1
14.7
22.2
26.6
19.8
19.4
91
Table 3.2. Continued. Percent of men with prevalent BPH by demographic and lifestyle
characteristics
Prevalent BPH
a
Total BPH
n=
%
15.2
20.7
31.8
57.8
28.1
25.9
31.0
37.3
=1,549
p o
O.0001
O.0001
Type
Treatment'
n=286
%
4.2
4.1
6.7
7.9
10.3
1.5
1.3
3.0
of BPH defining
Symptoms'
1
n=276
%
0.0
0.0
0.0
23.3
4.1
4.5
6.6
6.6
event
Physician
Diagnosis
6
n=987
%
11.0
16.6
25.1
26.6
13.7
16.9
20.1
27.6
IPSS baseline
1
0-3.5
3.6-6.5
6.6-14.0
14.1-19.9
Prostate volume (cm
3
)*
10.0-18.8 cm
3
18.9-25.1 cm
3
25.2-33.5 cm
3
33.6-100.0 cm
3
a
Prevalent BPH categorized hierarchically as surgical or medical treatment for BPH, an average IPSS > 14, or physician diagnosis
of BPH at baseline
b
P value from chi-square comparing total 'prevalent' BPH to no BPH
c
Surgical or drug treatment for BPH with or without physician diagnosis and 2 IPSS >14
d
Average IPSS >14 at baseline, with or without a history of physician diagnosis and no history of treatment
e
Physician diagnosis only (no symptoms or treatment)
f
<0.25 g/day(<l drink/month), 0.25-1.74 g/day(l-3 drinks/month), 1.75-12.9 g/day(l-6 drinks/week), 13.0-26.9 g/day(7-13
drinks/week), >27 g/day( >14 drinks/week)
8
Men with PSA>3.0 at baseline were excluded from participating in the PCPT
h
Self-report of father, brother or son having prostate cancer
1
Average IPSS from recruitment and baseline visits, men with IPSS at baseline > 20 were excluded from participating in the PCPT
' Excludes 626 men missing prostate volume and 76 men with prostate volume <10 or >100; prostate volume captured at the EOS
prostate biopsy (men without cancer) or diagnostic biopsy (men with cancer)
92
Table 3.3. Percent of men with
characteristics
Age
55-59
60-64
65-69
70+
Race
Caucasian
African American
Other
Smoking status
Current
Former/never
Body mass index (kg/m )
Normal (<25)
Overweight (25-29.9)
Obese (>30)
Physical activity
Sedentary
Light activity
Moderate activity
Very active
Alcohol (grams/day)
f
<0.25
0.25-1.74
1.75-12.9
13-26.9
>27
Baseline PSA (ng/ml)
8
0-1.0
1.1-2.0
2.1-4.0
Family history prostate cancer
11
Yes
No
prevalent plus incident BPH by demographic and
Prevalent plus incident BPH
ab
Total BPH
n=
%
36.3
43.4
54.2
58.5
45.6
39.0
51.8
41.6
45.9
46.0
46.2
43.6
48.0
45.6
43.7
48.6
50.9
47.0
43.4
43.2
40.6
40.3
47.8
55.6
46.2
45.5
=2,312
F
O.0001
0.07
0.12
0.32
0.11
O.0001
O.0001
0.69
lifestyle
Type of BPH defining event
Treatment
1
"
n=636
%
7.9
10.5
15.8
23.0
12.3
14.3
19.4
10.5
12.7
12.3
12.6
12.9
11.3
13.0
12.6
12.8
14.5
12.9
12.3
10.1
11.1
12.9
11.9
13.0
12.7
12.5
Symptoms'
n=605
%
11.1
12.3
12.6
12.0
11.9
9.1
16.5
13.7
11.8
10.7
12.6
12.3
14.2
11.5
11.3
12.2
13.4
11.2
11.9
12.0
9.4
11.1
12.0
14.1
11.9
11.9
c
Physician
Diagnosis'
1
n=l,071
%
17.3
20.6
25.8
23.6
21.5
15.6
15.9
17.4
21.4
23.1
21.0
18.4
22.5
21.1
19.8
23.7
23.0
22.8
19.1
21.0
20.8
16.4
23.9
28.1
21.6
21.0
93
Table 3.3. Continued. Percent of men with prevalent plus incident BPH by demographic and
lifestyle characteristics
IPSS baseline'
0-3.5
3.6-6.5
6.6-14.0
14.1-19.9
Prostate volume (cm
3
)*
10.0-18.8 cm
3
18.9-25.1 cm
3
25.2-33.5 cm
3
33.6-100.0 cm
3
Total BPH
n=
%
21.2
32.9
51.6
81.6
41.2
39.5
44.9
57.2
=2,312
P*
<0.0001
O.0001
Prevalent plus incident BPH
a
Type
Treatment
1
*
n=636
%
7.7
10.3
15.1
18.0
15.6
10.9
10.8
12.6
of BPH defining
Symptoms'
n=605
%
0.8
3.9
9.3
36.3
10.4
10.1
12.4
15.3
event
Physician
Diagnosis'
1
n=l,071
%
12.7
18.7
27.2
27.2
15.2
18.6
21.7
29.4
a
Prevalent plus incident BPH categorized hierarchically as surgical or medical treatment for BPH, an average IPSS > 14, or
physician diagnosis of BPH at baseline or during PCPT
b
Surgical or medical treatment for BPH with or without physician diagnosis and 2 IPSS >14
c
Average IPSS >14 at baseline, with or without a history of physician diagnosis and no history of treatment
d
Physician diagnosis only (no symptoms or treatment)
* P value from chi-square comparing total 'prevalent plus incident' BPH to no BPH
f
<0.25 g/day(<l drink/month), 0.25-1.74 g/day(l-3 drinks/month), 1.75-12.9 g/day(l-6 drinks/week), 13.0-26.9 g/day(7-13
drinks/week), >27 g/day( >14 drinks/week)
g
Men with PSA>3.0 at baseline were excluded from participating in the PCPT
h
Self-report of father, brother or son having prostate cancer
' Average IPSS from recruitment and baseline visits, men with IPSS at baseline > 20 were excluded from participating in the PCPT
J
Excludes 626 men missing prostate volume and 76 men with prostate volume <10 or >100
94
.2 3
2
©
p.
^
.iS
• a
«3
U
53
S5>
V
1 .
a
i
2
AH
«
Y
rl
V
1%
z
4>
c
a
n
s


CT.
NT>
NO
II
a
5
5
6

h
2
s
v
13
« m
l> O
!-
WD
!*>
> -
4>
S.
es
i -
M
O
S
01
w
S3
03
u
o
a.
I
CI
I
"O Ti -
es r~-
s* o
O C
o
H
i n
CS
CS
en
O
NO oo ON --<
en en NO f^
o oo
o
TI-
o
i n
o
Tl" 0 0
TT TT
rt O - H
t - TI- TI-
M H (J, O
i n TT i n T|-
OO ^t N rf *
i n i n en i n vo
<
0
.
0
0
0

O """!
cK ^
<
0
.
0
0
0
1

en NO
l-H CS
( N
Tr'
i n
en
ON
«~
oo
Tr'
0
.
0
0
0
3

rn
cs
<
0
.
0
0
0
1

Tr
cs
^
Tr'
p
ON
i n
T—t
_
TI-"
0
.
9

en
CS
0
.
8

ON
CS
t~;
CS
NO
CS
0
.
3
1

rt
en
0
.
5

CS
CS
o
en
ON
CS
CS
r
~*
c~-
cs
0
.
3
8

r—t
CS
0
.
1
5

o
CS
t-;
CO
,_
en
NO
en
ON
CS"
r^
©
en
^
0
.
7
2

CS
CS
0
.
8
3

ON
CS
ON
CS
TT
CS
r-;
CS
> n
CS
0 0
1—1
m
CS
NO
Tr
en
en
o
o
q h o o a
Os —i en in
« -i CS CS CS
CS
o
o
en
00
O
o
o
i n
en
O
cs t - Tr
CS CS •->
TT ON
CS CS
vq TT r-;
CS CS - n
o —< en >n
cs .-< en ©
CS CS CS CS
o
•—i ON NO OO TT
cs cs o <-< en
cs cs cs cs cs
O
.
0
0
0

Tr r-~
O en
cs cs
ON
NO
cs
oo
ON
CS
0
.
0
0
2

t
TT
cs
TT CS
TT OO
en ^H
0
.
8
1

r^
en
cs
cs
TT
cs
0
.
0
7

en
NO
cs
oo
en
cs
TI-
CS
cs
0
.
2
8

^
(
TI-
CS
NO
en
cs
0 0 —I
i n cs
cs cs
0
.
4
4

NO
TI-
CS
OO
TI-
CS
ON
CS
cs
0 0
en
cs
r^
NO
CS
95
I
«
w
".3
a.
cs
1.
61!
O
a
V
•8
Si
X)
> -
IU
o
s
cs
u
.2 s
SB CB
O W
« U
M ©
u
s
CS
u
s-
. P •=..,
ON
NO
NO
NO
m
i n
o
4>
O WO
"S H
ex
at
•a
6X
O
i n
CS
CS
•s a
o
o
o
o
V
""!
CS
ON
t~
•^r
i n
t—
o
o
o
p
00
CS
rr
c i
CS
o
ON
ci
oo
ci
cs
i n
i n
VO
CS
o
o
o
oo
NO
i—i
NO
CS
^r
ON
O
o
o
o
o
v
ci
o
o
o
VO
ci
m
T f
,
0
0
0
1

o
V
ON
O
T f
CS
ci
o
C I
o
r^
0 0
^o
I—1
oo
t—t
r—*
1 - H
r*-J
r—1
OO
1—1
1—t
• t
CS
r—<
c i
ci
* - H
o o n n ^
oo ON ON 2
CS
ON
ON
ON
t - r-
CS CS
> - ; ; t ~ ; • * f -
cs cs ci cs"
> -< CS I-H CI
CS CS CS CS
o
o
p
o
v
ON Tf O
TT i n Tt
o
o
o
o
V
o
o
p
d
v
c i
cs
o
in o cs oo
r- o
cs cs
CS O >—" C I
cs cs cs cs
• H - H ON t -
cs cs ~ ~
o
o
o
o
v
oo
cs
oo t— ro O
cs
o
O O O NO
i n oo oo
> —i CS CI
ON C>
cs cs
CI CS ^- NO
cs cs cs cs
CI CI *—' ON
CS CS CS •"-"
60
OH
i>
g
13
PQ
O
o
p p
o cs ci
.M
(Z)
IS)
£s
o O S c i
ta
«
i n
NO
3
.
6

4
.
0

'T'
6
.
6

o
o
cs
1
"fr
o
^-<
s
J3
o
>
<L>
r
o
s
t
a

P H
^^
S
o
0 0
oo
1
p
o
r*N
m pi H
a a s
o o 0
- *0 o
>n c i o
CS O <-i
• i i
ON CS NO
0 0 i/ S CI
<-* CS CI
S
1
1
op
c
a
1
w
i
t
h

c
e
r

c
a
s
e
s

t
a
t
e
c

r
o
s

o.
NO
e
l
u
d
e
s

4

X
n
c
e
r
s

c
1
^
+
7
(
3
-
G
l
e
a
s
o
n

•a
<
7
a

1
a
ja
I
n
c
l
u
d
e
s

G

^
;
e
r
s

u
+
OO
e
o
a
n
d

G
l
e
a
:

m
7
(
4
+

I
a
JJ
:
l
u
d
e
s

G

5
y
f
1 ,c
•a
<D
8
5
d
e
t

</^ *3"
>i
«
§•
S
>*
o
f
-
s
t
u
d

•a
s
i
e
c
t
e
d

"S
•o
o
r-.
D
R
E

r

a
b
n
o
r
m
a
l

n
d
/
o

S3
<
a
t
e
d

s
d

b
y

e
l
e
v

o.
p
r
o
m

c
o
•^
8
n
c
e
r

d
e
t

U
• 3
^~*
*e3
C
(D
^ O
• S
t 3
<D
O
(D
t5
e
r
s

o
C
rt
o
O
3
>%
*o
3
<4-
o
o
n

e
n
d
-
•8
s
«
c
e
r
s
d
'

s
6
4
4

D
R
E
(

n
o
r
m
a
l

o
<
C/3
(X
T3
«
e
l
e
v

r
o
m
p
t
e
d

b
y

n
o
t
p
:

§
•a
8
n
c
e
r

d
e
t
i

S3
•a
^ 6
r-
CN
Al
s~
S
~ t ^
•a
' C
•a
r*N
C--
d
a
y

ON
NO
(^1
O
m
2"
a>
d
r
i
n
k
s
/
w
e

NO
^,
g
/
d
a
>

ON
CS
VN
t - ^
n
t
h
)
,

1

n
k
s
/
m
o

d
r
i

(
1
-
3

>>
•a
~Sb
T f
-
1
.
7

m
t
h
)
,

0
.
2
5

o
r
i
n
k
/
i

T3
V,
^ ^
.
2
5

g
/
d
a
;

k
s
/
w
e
e
k

^ T3
u
2Z v
* -s
c
00
"5b
A
LU
S "
u
a
SI
u =
P H
«
C
1
i
•s
1

ffl
u.
03
C3
T
o
t

n
>/->
Os
re
m
>/-> ^ H
: '
Tt
Os
CO
=
P H
M
e
Z
%
so
,—
!
i—»
OS
o
*~-^
CN
O
I—1
oo
*"-;
>/ -f
t~-;
o
^J-
os
o
so
CO
^
ro"
Os
O
ro
*~~1
*—i
o
T3
O
SO
' 1
o
OS
o
v
-^ CN
O
' '
OS
•"I
>^r t -;
O
>/->
OS
o
t-~
CO
• - '
co"
OS
o
CO
'~~i
CN
a>
•O
o
s s
T f
»—1

CO
OS
o,
CO
p
•a
O
O
~
w
• o
»-.
a
o
o
u
1—t
a> aj
o ^3
a o
• *
^*

CN
OS
O^
CO
p
•a
O
O
~
J,
CN
0>
•o
O
3S S
o
H
-3-
r-~
oo
CN
^r
l - H
CN
SO
,
o
CN
CN
>/->
*""'
^
SO
T f
CO
oo
oo
so
CN
0 0
CN
r-
<D O
w> *•-
o 5
O
% o
w

g
r
a
d
e

o
-1
OS
"""l
f—1
OS
O
*—*
•3"
O
' '
CO
CN
SO"
t~-
o
t ^
Os
O
T.
^
-
CN"
OS
o
•sr
*"\
»-«
CD
-o
o
oo
*~*t
o
Os
O
^—' CO
p
1—1
CO
CN
so"
(^
o
t-~
OS
o
o
TT
i—t
^
Os
O
CO
*~i
CN
CD
T3
O
% s
t ^
^

CO
Os
o,
Tl -
o
T3
o
p
—»
y
^ H
CD
T3
O
so
^

CO
OS
O ,
Tl -
p
• o
O
O
•-<
y
CN
CD
T3
O
s s
O
t—
F^ • " " C
T^"
CN
o
CN
Os
»o
1—1
1—1
3
0
/

6
8
8

CN
O O
</} t-H
CO -S
o 5
-J O
% o
&>
•o
s
OS
I - H
t ^
OO
o
*s
CN
p
"
O
CN
oo"
SO
o
o
Os
O
>/->
CO
^
CN"
oo
o
i n
p
^
^
9^
CO
CO
t ^
• *
o
^-^ OS
r-;
o
o
0 0
•n"
CN
o
r-
SO
O
OO
CN
CN
ir\
V~l
o
CN
i—i
> —
CN
M
o
d
e
l

M
o
d
e
l

ITS
CN

CO
m
o
^-^ CN
OO
o
T3
o
p
^
u
• - H
(U
T3
O
OO
CN

m
•n
o
^—^ T f
0 0
o
T3
o
p
^
y
CN
IU
T3
O
s s
•>3-
r^
CO
i
r f
»-H
CN
CN
^^
O
CN
Os
v~t
'~
H
^
'3-
oo
'-
,
oo
oo
so
Cs|
CN
CO
« 1 .—1
<u o
V3 t-i
o 5
-J O
% o
so
o
3
r
-
c
a

o
so «rs
CN CN
v~> Tj -
0 0 0 0
o o
^—^ s—'
• * co
p p
r—t i—t
oo oo
• * . ^
>r? v-T
r~ t--
o o
SO SO
p p
^ ^ rt
O CN
</ -> ir>
—; ^ 4
rt" C N
0 0 0 0
o o
O -H
,—
!
,—
1
^ CN
a> D
•o -o
o o
2 S
^r <r>
CN CN
« OS Os
oo oo
2- ^
• O so
p p
T3 -a
o o
p p
r—1 *—1
u y
^ CN
a> a>
T3 T3
O O
s s
T f
r~-
CO
^
T—t
CN
o
1 (
o
CN
SO
V~l
l "
- 1
^
o
1—t
CN
0 0
oo
so
CN
Os
u-s
^r
( » ' —1
<U O
V3 H
o
% o
JS

s
C3
t
-
f
o
r
-
c

z
CN CN
CN CN
SO >/ ->
0 0 OO
o o
^^ ^-^ CN CN
O O
l—t I -H
OS Os
!—1 f - H
Os" o "
i n SO
o o
' t T f
OO OO
o o
t ^ t ^
i n i n
^-.' ^
o" Os"
os oo
o o
OS OO
I—1 T—t
^ H C N
(U 0)
T3 T3
O O
s s
OS Os
^ H »-H
„ „
r- t^-
0 0 OO
o o
CN CN
O O
13 T3
o o
p p
»—t 1-H
y u
^ H CN
<U <L>
-a is
o o
s s
f
i
n
i
i

<o
T3
SB
a .
03
1
O
X
1
111
3
s
! _ii
B g
03 g
3 ^
«2 W3
n a
O J3
5 "9 <2 -a
I
o
o ^
g CN
t " -a
•g Q Q
|1|
ill
Jit
[I!
ill
tSi
97
g. su
•a gv m
1\2~
eu -a i>
a
41
CM
I
e
Ml
g
"S
c
41
• a
=
PH
=
oi
a
s

>•
b *
PH
a
as.
'3
a
PQ
e
v
" 3
,g
"a
a
n
PH
PQ
o
Z,
vq
ci
u
«
m
O
o o
v o
VO
CN
CN
m
CN
T T
r ^
^r
CN
VO
»-H
»—1
m
r^
C^
' —i
VO
i n
CN
ON
o
CN
VO
vo
5 CZ)
0 ~ *
L> O
o l-
5 1
' o
t o
>H
41
a
n
c

w
t
a
l

o
H
CN
^
#, OO
oo
o
*—^
ON
o\
o
vo
p
t - - "
r^
o
*—'
o
ON
O
^_^
"Cf
^_
• * "
OO
o
*—
y
oo
ON
o
y
~*
n>
1 3
o
CN
^
r>
r~
oo
o
^-—'
oo
ON
©
f ~
O
(-»"
t~~
o
s
~-
/
*—1
ON
©
^^
•<r
^
• * "
o o
o
*"—•* oo
ON
o
a
<N
(D
1 3
O
s s
vo
o
,_;
r-"
oo
o
vo
ON
o
re
O
O
»—1
y
"—'
<L>
•o
o
VO
O
,_J
r-"
oo
o
VO
ON
o
re
O
o


CN
1)
-o
o
s s
o
oo
C I
CN
CN
O
CN
t ^
T f
ON
C N
i—i
i n
c^
CN
ON
•<t
CN
ON
O
CN
CN
OO
i n
~"~S W
CO i—I
CD O
CO ! - •
o 5
-T o ^ O
^
O
h-1
• * m
^ ^
n n
t~- VO
OO 0 0
o o
^—' ^—•
o o\
O ON
rt O
o o
^ ^
r^ oo"
t— t~-
S S
CS r o
ON ON
O O
^
vo~ ^o
T-H
1
"
- 1
rn" f^
S oo
d o,
» " oo
ON ON
o d
o
- H ( N
D U
T3 T3
O O
s s
00 OO
o o
r-J ^-H
P-" t~-"
00 OO
Q- 2-
t~~ t —
ON ON
d d
«! «
o o
p p
y u
** ( N
(D CD
T3 1 3
O O
s s
o
0 0
o
CN
CN
51
I—1
^ H
r^
^
VO
t—1
t —
T f
CN
ON
o
CN
0 0
- ^ w
CO i—i
U O
co H
a
a
- J O
* o
«u
•a
M
OX)
3
VO ON
r<-> c<N
n n
ON —<
•<3- i n
d d
* ^ > ^ ^-^
CN •<*•
OO OO
d d
CN Tj -
i n i n
r n " C<T"
• * -<t
d d
^—• ^ - ^
*—« i—t
0 0 OO
d d
x—*\ * —^
oo r o
i n v q
r n " - ^
i n i n
d d
^ —' s _ ^
CN CN
ON ON
d d
^ ^_
'—' CN
T3 1 3
o o
2 S
•<r t >
CN CN
»-H T-H
oo" ON"
i n i n
- d d
i n v o
o o o o
d d
re a
o o
p p
1—1 I—1
y w
r ^ CN
1) D
1 3 1 3
O O
s s
r~
•^f
CN
CN
^ 1 "
5
r~
i n
t—t
i n
^
ON
"cf
CN
CN
ON
o
CN
o
CN
- ^ co
CO ~ ^
D o
CO t-H
o 5
- J O
% o
so
cu
CO
3
«
1
O
fc.
OO 1 ^
»—1 »—1
n ^
ON ON
t — l ^ -
d d
* ~ ^ N^^/
r-~ vo
ON ON
d d
C"~ OO
p p
tr{ c-T
vq vq
d , d,
CN CN
oo o o
d d
^- ~v , —^
CN CN
t —i *—i
oo" oo"
VO VO
d d
^—'
x
—'
(~~ f~-
OO 0 0
d d
•3 W
^ H CN
<D <U
1 3 1 3
o o
s s
v o i n
o o
»—4 i—*
t - ^ r-"
r- r~
d d
o o
ON ON
d d
re re
O O
o p
*—t 1-H
„ y
^ H CN
U D
1 3 1 3
o o
£ S
O
OO
T f
• ^ j -
CN
5
VO
r-
c^
T f
CN
ON
•—I
i n
j—t
CN
i—i
CN
CN
ON
o
CN
i n
CN
- ^ co
M ^ H
«J O
co H
o 5
- J O
* o
CO
S
A
1
u
o
< * -
1
4-*
e
Z
rt O
CN CN
n n
T f T f
OO 0 0
d d
^ - ^ *~~s
»—t T-H
p p
> - ! ^ - 1
ON O
"1 ^
i n " vo"
f~; t - ;
2- 2-
i n i n
ON ON
d d
^ ^^
CN CN
CN cr\
vo" vo"
OO OO
d d
^—^ ^—^
(~~ ( ^
p p
^ H ^
•3 «
^ H CN
(U <D
1 3 1 3
o o
2 2
VO VO
r—( .—(
I—1 »—1
t - - " t --"
oo oo
d d
^ —' N—'
^ H ^ H
O O
— ^
re re
o o
o o
~ ~
•3 W
^ CN
a> a>
1 3 1 3
O O
S 2
c SI ^
-o
1
•s
:
e
(
w
h
i
t
e
,

b
l
a
c
k

2
l
i
n
e
a
r
)
,

t r
_c
i
a
s
e
]

a
g
e

a
t

b

£ < 4 - l
I
j
U
!

VJ
<
l
e
i

1

o
B
P
H

C4—<
o
y
p
e

J3
O
s <4-i
o
c
d

f
o
r

a
s
s
o
c
i
a
t
i
i

9
a
l
B
P
H

o
< 4 - l
o
c
:
s
o
c
i
a
t
i
o

i _
<a
n
o
.

r
e
n
c

'
R
e
f
e

i
n
d
e
;

CA
s
T3
O
. O
J 3
O
e
(
w
h
i
t
e
,

b
l
a
c
k

r
a
c

C
e
(
l

.c
a
s
e
l

a
g
e

a
t

b

f
o
r

t
e
d

j
u
s

• o
<
CN
*H3
M
o
d

o
d
e
l

w
a
s

u
s
e
d

P
o
i
s
s
o
n

m

a?
go
n
v
e

d

n
o
t

c
o

•3
CO
1

m
o
d
'

cct
1
L
o
g
-
b
i
n
o

sa
o
r

a
b
n
o
r
m
a
l

D

P
S
A
a
i

TB
^
e
l
e
v

p
t
e
d

b
y

s o
L~
o.
o
"o
e
t
e

e
r
d

C
a
n
e

A

o
r

a
b
n
o
r
m
a
l

D
]

a
t
e
d
P
S

>
*v
> 1
r
o
m
p
t
e
d

o.
o
s
c
o
o
e
t
e

e
r
d

'
C
a
n
e

98
S3
!S
a
a
u
CD
•o
a
u
6JD
J3
b
V
u
S3
«
w
K
o
1-
o,
«
*-
W
P H
CQ
-W
S3
V
•a
>
P
r
e

1.
a.
a
o
i-
PH
a
WD
S3
o>
P3
S3

S3

o °°
s-
CO g
H 3
* i
g
>
QJ
ex
.S
S3
d
e
f
i

B
P
H

<«-!
o
1 1
a.
>>
H
c
i
a
n

-
K
>?
JS
0 -
• c
;
n
o
s
i
s

5
%
C

« w
•° ai
3
^ ~ s
CB >—1
t
o
r
n
;

%
C

a , wi
t/ 2 pd
h H
55
u
V «
S s?
•w i n
« ON
H
3
o
<N
r f
CN
OS
o
CN
SO
^ H
o\
oo
o
CN
O
.-«
co
CN
7
,
1
.

t-~
7
(
0
.

o
o
Os
e n
i—i
» - H
Os
O ,
CN
^
SO
^ H
1—1
oT
0 0
o
Cs|
p
.-!
CO
CN
f -"
f -
8
(
0
.

OS
O
Os
co
I—1
i—1
Os
O ,
CN
*—i
O
t ^
C N
• *
CN
o
CN
SO
O l >
oo oo
o" CN"
V~t l O
o d
HO OS
Os OS
o o
, ^ ^ y - ^ .
1— oo
OS Os
j—1 i—1
o o
d d
Os Os
CN CN
d d
*—N / —^
TJ- SO
OO OS
CN CN
«"T co"
• ^" ITS
d d
o >o
<N <N
a
pa
o
Z
, -H CN
o o
=
P H
pa
T
o
t
a

)—1
O
s?
no
Os^
W.
>/ o
</ o
^^
m
^ H
CN"
Os
O^
r n
o
i - H
of
Os
o ,
CN
o
oo
oo
so
CN
o
SO
o o
o o
oo
oo
so
CN
CO
o
CO
i n
>-•
CN"
l/->
d
Os
OO
oo
"1
'-'
I D
d
<N
Os
o
o
V
S3
o
u
s
o
1 .
w
S3
<3
U
o
%
o
n
t
i

o
u
<u
• O
C3
'-x
^
©
-1
_
i
t )
T3
O
s
° o
to
O
6X1 S
&X
•M
oa .S
° S'
M. S'
s 8:
IJ! .
S 03
c —
o--S
•a
c
•°.2 .
+
• a
U t N
f r - S ^
•a
G
as
0 ^ o
i G
§
-3
•o
+ 2
U —i
( N S C
• s i ?
99
w
•~
a>
u
a
a
w
V
13
«
b
M
A
W>
3
^
w
b
l*
4>
U
S3
« U
41
*- cs
• * - »
«>
o
a
X
• * *
?
rt
*6ft
C3
» - H
a.
a.
^
X
u • • 4
-*-
-** VI
o
h
PH
s
£f
"3 0)
pa
• 4 - *
8
4>
T3
• • 4
w
.S3

_s
"a.
**
s
&)
93
>• 4>
U
n.
< * -
o 5A
s
o
i
a
t

U
o
Vi
<
GO
c$

3
03 H
+
so
•a
a
s.
sx
a
o t»
es
&>
3
6A
CS
•a
si
a>
•O
B
PM
PQ
- W
•O
• p *
u
.a
6 A
_s
"a
+-*
S3
a>
C3
>
i-
P-
£•-3
§ 0
w
PH
PQ
"3
T
o
t

u
0
s
-
U~l
o^
"(4
i—t
>o
t ^
^
ON
o
n
PM
PQ
o
Z,
1
7
5
1

3
0
/

oo
ON"
o
NO
ON
.
6
6
,

1

o
/
8
0
5

•<r
ON
,
5
5
,
1

o
o
<N
©
,
5
7
,
2
.

o
oo
o
,_,
1
7
5

3
0
/

oo
T —1
ON"
o ,
NO
ON
i —i
.
6
6
,

o,
ON
4
7
2

i/ ~>
NO
, —«
o,
NO
NO
NO
^ H
NO"
©^
IT)
NO
OO <N
«n
©
ro
ON" NO"
IT) NO
©' ©
5
^*
i/->" v>"
o o
NO r~
ON ON
o o
ON
©
oo
o
NO
03
e
° o
o o
oo oo
NO NO
ON ON
o o
o o
i n
o
ON
o
(N
ON ON
O O
•a -a
o o
OO <N
ON NO
IT) NO
NO r-
ON ON
o o
o o
o o
(3
o
1
5
o
o
£ ~o *o
1 o"§
60
!
o
^4
T3
rr
Q-
03
> r
m
' u
s
M
t i l
*Sil
e
r
v
a

c
e

I
n
t

•a
>
n
f
i

^
t f
<N
g
r
a
d
e

c
o
.2
•8
a
A
d
j
u
;


1
5
2 u a s
s
a>
oi
T3
<
(N
-n
o
S
f l
- 9
100
References
1. Bosch JL, Hop WC, Kirkels WJ, Schroder FH. Natural history of benign
prostatic hyperplasia: appropriate case definition and estimation of its
prevalence in the community. Urology 1995;46(3):34-40.
2. Guess HA. Epidemiology and natural history of benign prostatic hyperplasia.
Urol Clin North Am 1995;22:247-261.
3. Kirby RS. The natural history of benign prostatic hyperplasia: what have we
learned in the last decade? Urology 2000;56(5):3-6.
4. Sidney S, Quesenberry CP, Sadler MC, Guess HA, Lydick EG, Cattolica EV.
Incidence of surgically treated benign prostatic hypertrophy and of prostate
cancer among blacks and whites in a prepaid health care plan. Am J Epidemiol
1991;134:825-829.
5. Foster CS. Pathology of benign prostatic hyperplasia. Prostate 2000;45(S9):4-
14.
6. Barry MJ, Floyd JF, O'Leary MP, Bruskewitiz RC, Holtgrewe HL, Mebust
WK, Cockett AT. The American Urological Association symptom index for
benign prostatic hyperplasia. J Urol 1992;148(5):1549-1557.
7. Girman CJ, Epstein RS, Jacobsen SJ, Guess HA, Panser LA, Osterling JE,
Lieber MM. Natural history of prostatism: Impact of urinary symptoms on
quality of life in 2115 randomly selected community men. Urology
1994;44(6):825-831.
8. Garraway M, Kriby RS. Benign prostatic hyperplasia: Effects on quality of life
and impact on treatment decisions. Urology 1994;44(5):639-636.
9. Garraway WM, Russell EBA, Lee RJ, Collins GN, McKelvie GB, Hehir M,
Rodgers ACN, Simpson RJ. Impact of previously unrecognized benign
prostatic hyperplasia on the daily activities of middle-aged and elderly men.
British Journal of General Practice 1993;43:318-321.
10. Arrighi HM, Guess HA, Metter EJ, Fozard JL. Symptoms and signs of
prostatism as risk factors for prostatectomy. The Prostate 1990;16(3):253-261.
11. Guess HA. Benign prostatic hyperplasia and prostate cancer. Epidemiol Rev
2001;23:152-158.
101
12. Kristal AR, Arnold KB, Schenk JM, Neuhouser ML, Weiss N, Goodman PJ,
Antvelink CM, Penson DF, Thompson IM. Race/ethnicity, obesity, health
related behaviors and the risk of symptomatic benign prostatic hyperplasia:
results from the Prostate Cancer Prevention Trial. J Urol 2007; 177(4): 1395-
1400.
13. Giovannucci E, Rimm EB, Chute CG, Kawachi I, Colditz GA, Stampfer MJ,
Willett WC. Obesity and benign prostatic hyperplasia. Am J Epidemiol
1994;140(11):989-1002.
14. Meigs JB, Mohr B, Barry MJ, Collins MM, McKinlay JB. Risk factors for
clinical benign prostatic hyperplasia in a community-based population of
healthy aging men. J Clin Epidemiol 2001;54:935-944.
15. Chokkalingam AP, Gao YT, Deng J, Stanczyk FZ, Sesterhenn IA, Mostofi FK,
Fraumeni JF, Hsing AW. Insulin-like growth factors and risk of benign
prostatic hyperplasia. Prostate 2002;52:98-105.
16. Morrison AS. Risk factors for surgery for prostatic hypertrophy. Am J
Epidemiol 1992;135:974-980.
17. Gann PH, Hennekens CH, Longcope C, Verhoek-Oftedahl W, Grodstein F,
Stampfer MJ. A prospective study of plasma hormone levels, nonhormonal
factors and development of benign prostatic hyperplasia. Prostate 1995;26:40-
49.
18. Lee SJ, Min HG, Choi SH, Kim YJ, Oh SW, Kim YJ, Park Y, Kim SS. Central
obesity as a risk factor for prostatic hyperplasia. Obesity 2006;14:172-179.
19. Hammarsten J, Hogstedt B, Holthuis N, Mellstrom D. Components of the
metabolic syndrome-risk factors for the development of benign prostatic
hyperplasia. Prostate Cancer Prostatic Dis 1998;1(3):157-162.
20. Parsons JK, Carter HB, Partin AW, Windham BG, Metter EJ, Ferrucci L,
Landis P, Platz EA. Metabolic factors associated with benign prostatic
hyperplasia. J Clin Endocrinol Metab 2006;91(7):2562-2568.
21. Roehrborn CG, Schulman C, Arrosagarai P, Duggan A, Major-Walker K,
Morrill B. Baseline data from the combination of Avodart™ and tamsulosin
(combat) trial: relationships between body mass index (bmi) and luts/bph
measures. Eur Urol 2006;5(2 Suppl 195):692.
102
22. Haque WA, Garg A. Adipocyte biology and adipocytokines. Clin Lab Med
2004;24(l):217-234.
23. Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB. Elevated C-
reactive protein levels in overweight and obese adults. JAMA
1999;282(22):2131-2135.
24. Yalow RS, Glick SM, Roth J, Berson SA. Plasma insulin and growth hormone
levels in obesity and diabetes. Ann New York Acad Sci 1965;131:357-373.
25. Considine RV, Sinha MK, Herman ML, Kriauciunas A, Stephens TW, Nyce
MR, Ohannesian JP, Marco CC, McKee LJ, Bauer TL, et al. Serum
immunoreactive-leptin concentrations in normal-weight and obese humans. N
Engl J Med 1996;334(5):292-295.
26. Trayhurn P, Wood IS. Adipokines: inflammation and the pleiotropic role of
white adipose tissue. Br J Nutr 2004;92(3):347-355.
27. Arita Y, Kihara S, Ouchi N, Takahashi M, Maeda K, Miyagawa J, Hotta K,
Shimomura I, Nakamura T, Miyaoka K. Paradoxical decrease of an adipose-
specific protein, adiponectin, in obesity. Biochem Biophys Res Commun
1999;257(l):79-83.
28. Arita Y, Kihara S, Ouchi N, Takahashi M, Maeda K, Miyagawa J, JHotta K,
Shimomura I, Nakamura T, Miyaoka K, Kuriyama H, Nishida M, Yamashita S,
Okubo K, Matsubara K, Muraguchi M, Ohmoto Y, Funahashi T, Matsuzawa Y.
Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity.
Biochem Biophys Res Commun 1999;257:79-83.
29. Kelesidis I, Kelesidis T, Mantzoros CS. Adiponectin and cancer: a systematic
review. Br J Cancer 2006;94:1221-1225.
30. Yamauchi T, Kamon J, Minokoshi Y, Ito Y, Waki H, Uchida S, Yamashita S,
Noda M, Kita S, Ueki K, Eto K, Akanuma Y, Froguel P, Foufelle F, Ferre P,
Carling D, Kimura S, Nagai R, Kahn BB, Kadowaki T. Adiponectin stimulates
glucose utilization and fatt-acid oxidation by activating AMP-activated protein
kinase. Nature Medicine 2002;8(11): 1288-1295.
31. Mistry T, Digby JE, Chen J, Desai KM, Randeva HS. The regulation of
adiponectin receptors in human prostate cancer cell lines. Biochem Biophys
Res Commun 2006;348:832-838.
103
32. Bub JD, Myiyazaki T, Iwamoto Y. Adiponectin as a growth inhibitor in
prostate cancer cells. Biochem Biophys Res Commun 2006;340:1158-1166.
33. Combs T, Berg A, Obici S, Scherer PE, Rossetti L. Endogenous glucose
production is inhibited by the adipose-derived protein Acrp30. Journal of
Clinical Investigation 2001;108(12):1875-1881.
34. Berg AH, Combs TP, Du X, Brownlee M, Scherer PE. The adipocyte-secreted
protein Acrp30 enhances hepatic insulin action. Nature Medicine
2001;7(8):947-953.
35. Duncan BB, Schmidt MI, Pankow JS, Bang H, Couper D, Ballantyne CM,
Hoogeveen RC, Heiss G. Adiponectin and the development of type 2 diabetes.
The Atherosclerosis Risk in Communities Study. Diabetes 2004;53:2473-2478.
36. Lu JY, Huang KC, Chang LC, Huang YS, Chi YC, Su TC, Chen CL, Yang WS.
Adiponectin: a biomarker of obesity-induced insulin resistance in adipose
tissue and beyond. J Biomed Sci 2008;15(5):565-576.
37. Ryo M, Nakamura T, Kihara S, Kumada M, Shibazaki S, Takahashi M, Nagai
M, Matsuzawa Y, Funahashi T. Adiponectin as a biomarker of the metabolic
syndrome. Circ J 2004;68:975-981.
38. Michalakis K, Williams CJ, Mitsiades N, Blakeman J, Balafouta-Tselenis S,
Giannopoulous A, Mantzoros CS. Serum adiponectin concentrations and tissue
expression of adiponectin receptors are reduced in patients with prostate cancer:
a case control study. Cancer Epidemiol Biomarkers Prev 2007;16(2):308-313.
39. Gotkas S, Yilmaz MI, Caglar K, Sonmez A, Kilic S, Bedir S. Prostate cancer
and adiponectin. Urology 2005;65:1168-1172.
40. Nazian SJ, Cameron DF. Temporal relation between leptin and various indices
of sexual maturation in the male rat. J Androl 1999;20(4):487.
41. Cioffi J A, Shafer AW, Zupancic TJ, Smith-Gbur J, Mikhail A, Platika D,
Snodgrass HR. Novel B 219/OB receptor isoforms: Possible role of leptin in
hematopoiesis and reproduction. Nature Medicine 1996;2(5):585-589.
42. Onuma M, Bub J, Rummel T, Iwamoto Y. Prostate cancer cell-adipocyte
interaction: leptin mediates androgen-independent prostate cancer cell
proliferation through c-Jun NH2-terminal kinase. J Biol Chem 2003;278:42660-
42667.
104
43. Haynes WG, Sivitz WI, Morgan DA, Walsh SA, Mark AL. Sympathetic and
cardiorenal actions of leptin. Hypertension 1997;30(3 Pt 2):619-623.
44. McVary KT, Razzaq A, Lee C, Venegas MF, Rademaker A, McKenna KE.
Growth of the rat prostate gland is facilitated by the autonomic nervous system.
Biology of Reproduction 1994;51(1):99-107.
45. Glynn RJ, Champion EW, Bouchard GR, Silbert JE. The development of
benign prostatic hyperplasia among volunteers in the Normative Aging Study.
Am J Epidemiol 1985;121:71-90.
46. Hammarsten J, Hogstedt B. Clinical, anthropometric, metabolic and insulin
profile of men with fast annual growth rates of benign prostatic hyperplasia.
Blood Pressure 1999;8:29-36.
47. McVary KT, Rademaker A, Lloyd GL, Gann P. Autonomic nervous system
overactivity in men with lower urinary tract symptoms secondary to benign
prostatic hyperplasia. J Urol 2005;174(4):1327-1333.
48. Emilsson V, Liu YL, Cawthorne MA, Morton NM, Davenport M. Expression
of the functional leptin recepton mRNA in pancreatic islets and direct inhibitory
action of leptin on insulin secretion. Diabetes 1997;46(2):313-317.
49. Sivitz WI, Walsh S A, Morgan DA, Thomas MJ, Haynes WG. Effects of leptin
on insulin sensitivity in normal rats. Endocrinology 1997;138:3395-3401.
50. Mark AL, Correia ML, Rehmouni K, Haynes WG. Selective leptin resistance:
a new concept in leptin physiology with cardiovascular implications. Journal of
Hypertension 2002 ;20.
51. Rohrmann S, Platz EA, Smit E, Giovannucci E. Re: Body size and serum levels
of insulin and leptin in relation to the risk of benign prostatic hyperplasia. J
Urol 2004; 172(2):779.
52. Lagiou P, Signorello LB, Trichopoulos D, Tzonou A, Trichopoulou A,
Mantzoros CS. Leptin in relation to prostate cancer and benign prostatic
hyperplasia. Int J Cancer 1998;76:25-28.
53. Saglam K, Aydur E, Yilmaz M, Goktas S. Leptin influences cellular
differentiation and progression in prostate cancer. J Urol 2003; 169(4): 1308-
1311.
105
54. Dahle SE, Chokkalingam AP, Gao YT, Deng J, Stanczyk FZ, Hsing AW. Body
size and serum levels of insulin and leptin in relation to the risk of benign
prostatic hyperplasia. J Urol 2002;168(2):599-604.
55. Hovorka R, Jones RH. How to measure insulin secretion. Diabetes Metab Rev
1994;10(2):91-117.
56. Bonora E, Zavaroni I, Bruschi F, Alpi F, Pezzarossa A, Guerra L, Dall'Aglio E,
Coscelli C, Butturini U. Peripheral hyperinsulinemia of simple obesity:
pancreatic hypersecretion or imparled insulin metabolism? J Clin Endocrinol
Metab 1984;59(6):1121-1127.
57. Abate N. Insulin resistance and obesity. The role of fat distribution pattern.
Diabetes Care 1996;19:292-294.
58. Carmena MJ, Fernandez-Moreno MD, Prieto JC. Characterization of insulin
receptors in isolated epithelial cells of rat ventral prostate: effect of fasting. Cell
Biochem Funct 1986;4(l):19-24.
59. Rohrmann S, Smit E, Giovannucci E, Platz EA. Association between markers
of the metabolic syndrome and lower urinary tract symptoms in the Third
National Health and Nutrition Examination Survey (NHANES III). Int J
Obesity 2005;29:310-316.
60. Nandeesha H, Koner BC, Dorairajan LN, Sen SK. Hyperinsulinemia and
dyslipidemia in non-diabetic benign prostatic hyperplasia. Clinica Chimica
Acta2006;370(l-2):89-93.
61. Hammarsten J, Hogstedt B. Hyperinsulinaemia as a risk factor for developing
benign prostatic hyperplasia. Eur Urol 2001;39:151-158.
62. Di Silverio F, Gentile V, De Matteis A, Mariotti G, Giuseppe V, Luigi PA,
Sciarra A. Distribution of inflammation, pre-malignant lesions, incidental
carcinoma in histologically confirmed benign prostatic hyperplasia: A
retrospective analysis. Eur Urol 2003;43(2): 164-175.
63. Roehrborn CG, Kaplan SA, Noble WD, Lucia MS, Slawin KM, McVary K,
Kusek JW, Nyberg LM. The impact of acute or chronic inflammation in
baseline biopsy on the risk of clinical progression of BPH: results from the
MTOPS study. J Urol 2005;173(S4):346.
106
64. Nickel JC, Downey J, Young I, Boag S. Asymptomatic inflammation and/or
infection in benign prostatic hyperplasia. BJU Int 1999;84(9):976-981.
65. Gerstenbluth RE, Seftel AD, Maclennan GT, Nagarjun Rao R, Corty EW,
Ferguson K, Resnick MI. Distribution of chronic prostatitis in radical
prostatectomy specimens with up-regulation of BCL-2 in areas of
inflammation. J Urol 2002;167(5):2267-2270.
66. Nickel JC, Roehrborn CG, O'Leary MP, Bostwick DG, Somerville MC,
Rittmaster RS. The relationship between prostate inflammation and lower
urinary tract symptoms: Examination of baseline data from the REDUCE trial.
Eur Urol 2008;54(6):1379-1384.
67. Kefi A, Koseoglu H, Celebi I, Yorukoglu K, Esen A. Relation between acute
urinary retention, chronic prostatic inflammation and accompanying elevated
prostate-specific antigen. Scandinavian Journal of Urology and Nephrology
2006;40(2): 155-160.
68. Mishra VC, Allen DJ, Nicolaou C, Sharif H, Hudd C, Karim OMA, Motiwala
HG, Laniado ME. Does intraprostatic inflammation have a role in the
pathogenesis and progression of benign prostatic hyperplasia. British Journal of
Urology International 2007;100(2):327-331.
69. Feneley MR, Span PN, Schalken JA, Harper M, Griffiths K, Holmes K, Kirby
RS. A prospective randomized trial evaluating tissue effects of finasteride
therapy in benign prostatic hyperplasia. Prostate Cancer Prostatic Dis
1999;2(5/6):277-281.
70. Kramer G, Steiner GE, Handisurya A, Stix U, Haitel A, Knerer B, Gessl A, Lee
C, Marberger M. Increased expression of lymphocyte-derived cytokines in
benign hyperplastic prostate tissue, identification of the producing cell types,
and effect of differentially expressed cytokines on stromal cell proliferation.
Prostate 2002;52(l):43-58.
71. Steiner GE, Newman ME, Paikl D, Stix U, Memaran-Dagda N, Lee C,
Margberger MJ. Expression and Function of Pro-Inflammatory Interleukin IL-
17 and IL-17 Receptor in Normal, Benign Hyperplastic, and Maliganant
Prostate. The Prostate 2003;56(3):171-182.
107
72. Steiner GE, Stix U, Handisurya A, Willheim M, Haitel A, Reithmayr F, Paikl
D, Ecker RC, Hrachowitz K, Kramer G, Lee C, Marberger M. Cytokine
expression pattern in benign prostatic hyperplasia infiltrating T cells and impact
of lymphocytic infiltration on cytokine mRNA profile in prostatic tissue. Lab
Invest 2003;83(8):1131-1146.
73. Handisurya A, Steiner GE, Stix U, Ecker RC, Pfaffeneder-Mantai S, Langer D,
Kramer G, Memaran-Dadgar N, Marberger M. Differential expression of
interleukin-15, a pro-inflammatory cytokine and T-cell growth factor, and its
receptor in human prostate. Prostate 2001 ;49(4):251-262.
74. Kessler OJ, Keisari Y, Servadio C, Abramovici A. Role of chronic
inflammation in the promotion of prostatic hyperplasia in rats. J Urol
1998;159(3):1049-1053.
75. St. Sauver JL, Jacobsen DJ, McGree ME, Lieber MM, Jacobsen SJ. Protective
association between nonsteroidal antiinflammatory drug use and measures of
benign prostatic hyperplasia. Am J Epidemiol 2006;164(8):760-768.
76. Elkahwaji JE, Zhong W, Jopkins WJ, Bushman W. Chronic bacterial infection
and inflammation incite reactive hyperplasia in a mouse model of chronic
prostatitis. Prostate 2007;67(1): 14-21.
77. Kirby RS, Lowe D, Bultitude MI, Shuttleworth KED. Intra-prostatic urinary
reflux: an aetiological factor in abacterial prostatitis. British Journal of Urology
International 1982;54(6):729-731.
78. De Marzo AM, Platz EA, Sutcliffe S, Xu J, Gronberg H, Drake CG, Nakai Y,
Isaacs WB, Nelson WG. Inflammation in prostate carcinogenesis. Nature
Reviews Cancer 2007;7(4):256-269.
79. Rivero VE, Motrich RD, Maccioni M, Riera CM. Autoimmune etiology in
chronic prostatitis syndrome: an advance in the understanding of this
pathology. Critical Reviews in Immunology 2007;27(l):33-46.
80. Drachenberg DE, Elgamal AAA, Rowbotham R, Peterson M, Murphy GP.
Circulating levels of interleukin-6 in patients with hormone refractory prostate
cancer. Prostate 1999;41(2):127-133.
81. Rohrmann S, De Marzo AM, Smit E, Giovannucci E, Platz EA. Serum C-
reactive protein concentration and lower urinary tract symptoms in older men in
108
the Third National Health and Nutrition Examination Survey (NHANES III).
Prostate 2005;62(l):27-33.
82. Turner RD, Belt E. The results of 1,694 consecutive simple perineal
prostatectomies. J Urol 1957;77(6):853-863.
83. Sommers SC. Endocrine changes with prostatic carcinoma. Cancer
1957;10:345-358.
84. Labess M. Occult carcinoma in clinically benign hypertropy of the prostate; a
pathological and clinical study. J Urol 1952;68(6):893-896.
85. McNeal JE, Redwine EA, Freiha FS, Stamey TA. Zonal distribution of prostatic
adenocarcinoma correlation with histologic pattern and direction of spread. The
American Journal of Surgical Pathology 1988;12(12):897-906.
86. Schroder FH, Blom JH. Natural history of benign prostatic hyperplasia (BPH).
The Prostate 1989; 15(52): 17-22.
87. De Marzo AM, Coffey DS, Nelson WG. New concepts in tissue specificity for
prostate cancer and benign prostatic hyperplasia. Urol Int 1999;53(Suppl
3a):29-39.
88. Bostwick DG, Cooner WH, Denis L, Jones GW, Scardino PT, Murphy GP. The
association of benign prostatic hyperplasia and cancer of the prostate. Cancer
1992;70(1):291-301.
89. Ornstein DK, Rao GS, Smith DS, Andriole GL. The impact of systematic
prostate biopsy on prostate cancer incidence in men with symptomatic benign
prostatic hyperplasia undergoing transuretheral resection of the prostate. J Urol
1997;157:880-884.
90. Lee C. Role of androgen in prostate growth and regression: stromal-epithelial
interaction. The Prostate 1996;6:52-56.
91. McConnell J, Bruskewitz R, Walsh P, Andriole GL, Lieber M, Holtgrewe HL,
Albertsen P, Roehrborn CG, Nickel JC, Wang DZ, Taylor AM, Waldstreicher J.
The effect of finasteride on the risk of acute urinary retention and the need for
surgical treatment among men with benign prostate hyperplasia. N Engl J Med
1998;338:557-563.
92. Thompson IM, Goodman PJ, Tangen CM, Lucia MS, Miller GJ, Ford LG,
109
Lieber MM, Cespedes RD, Atkins JN, Lippman SM, Carlin SM, Rayan A,
Szczepanek CM, Crowley JJ, Coltman CA. The influence of finasteride on the
development of prostate cancer. N Engl J Med 2003;349(3):215-224.
93. Shah US, Getzenberg RH. Fingerprinting the diseased prostate: associations
between BPH and prostate cancer. J Cell Biochem 2004;91:161-169.
94. Schenk JM, Kristal AR, Neuhauser ML, Tangen CM, White E, Lin DW, Kratz
M, Thompson IM. Biomarkers of systemic inflammation and risk of incident,
symptomatic benign prostatic hyperplasia: results from the Prostate Cancer
Prevention Trial. Am J Epidemiol 2010;171(5):571-582.
95. Armenian HK, Lilienfeld AM, Diamond EL, Bross ID. Relation between
benign prostatic hyperplasia and cancer of the prostate. Lancet 1974;2(115-17).
96. Chokkalingam AP, Nyren O, Johansson JE, Gridley G, McLaughlin JK, Adami
HO, Hsing AW. Prostate carcinoma risk subsequent to diagnosis of benign
prostatic hyperplasia. Cancer 2003;98(8):1727-1734.
97. Hiatt RA, Armstrong MA, Klatsky AL, Sidney S. Alcohol consumption,
smoking, and other risk factors and prostate cancer in a large health plan cohort
in Californa (United States). Cancer Causes Control 1994;5:66-72.
98. Simons BD, Morrison AS, Young RH, Verhoek-Oftedahl W. The relation of
surgery for prostatic hypertropy to carcinoma of the prostaet. Am J Epidemiol
1993;138(5):294-300.
99. Greenwald P, Kirmss V, Polan AK, Dick VS. Cancer of the prostate among
men with benign prostatic hyperplasia. J Natl Cancer Inst 1974;53(2):335-340.
100. Jacobsen SJ, Girman CJ, Guess HA, Chute CG, Oesterling JE, Lieber MM. Do
prostate size and urinary flow rates predict health care-seeking behavior for
urinary symptoms in men? Urology 1995;45(l):64-69.
101. Sheldon CA, Williams RD, Fraley EE. Incidental carcinoma of the prostate: a
review of the literature and critical reappraisal of classification. J Urol
1980;124(5):626-631.
102. Meigs JB, Barry MJ, Oesterling JE, Jacobsen SJ. Interpreting results of
prostate-specific antigen testing for early detection of prostate cancer. Journal
of General Internal Medicine 1996;11(9):505-512.
110
103. Uzzo RG, Wei JT, Waldbaum RS, Perlmutter AP, Byrne JC, Vaughan ED. The
influence of prostate size on cancer detection. Urology 1995;46(6):831-836.
104. Platz EA, Smit E, Curhan GC, Nyberg LM, Giovannucci EL. Prevalence of and
racial/ethnic variation in lower urinary tract symptoms and noncancer prostate
surgery in U.S. men. Urology 2002;59(6):877-883.
105. Neuhouser ML, Schenk JM, Song YJ, Tangen CM, Goodman PJ, Pollak M,
Penson DF, Thompson IM, Kristal AR. Insulin-like growth factor-1, insulin-
like growth factor binding protein-3 and risk of benign prostate hyperplasia in
the Prostate Cancer Prevention Trial. Prostate 2008;68(13):1477-1486.
106. Kristal AR, Schenk JM, Song YJ, Arnold KB, Neuhouser ML, Goodman PJ,
Lin D, Stanczyk F, Thompson IM. Serum steroid and sex hormone binding
globulin concentrations and the risk of incident benign prostatic hyperplasia
(BPH): Results from the Prostate Cancer Prevention Trial. Am J Epidemiol
2008;168(12):1416-1424.
107. Kristal AR, Arnold KB, Schenk JM, Neuhouser ML, Goodman PJ, Penson DF,
Thompson IM. Dietary patterns, supplement use and risk of symptomatic
benign prostatic hyperplasia (BPH): Results from the Prostate Cancer
Prevention Trial. Am J Epidemiol 2008;167(8):925-934.
108. Breslow NE, Day NE. Statistical methods in cancer research. Vol 1 - The
analysis of case-control studies. Davis W, editor. Lyon: Intl Agency for
Research on Cancer; 1980.
109. Richter EA, Mikines KJ, Galbo J, Kiens B. Effect of exercise on insulin action
in human skeletal muscle. J Appl Physiol 1989;66:876-885.
110. Bluher M, Bullen JW, Lee JH, Kralisch S, Fasshauer M, Kloting N, Niebauer J,
Schon MR, Williams CJ, Mantzoros CS. Circulating adiponectin and
expression of adiponectin receptors in human skeletal muscle: Associations
with metabolic parameters and insulin resistance and regulation by physical
training. J Clin Endocrinol Metab 2006;91:2310-2316.
111. Barry MJ, Cockett AT, Holtgrewe HL, McConnell JD, Sihelnik SA, Winfield
HN. Relationship of symptoms of prostatatism to commonly used physiological
and anatomical measures of the severity of benign prostatic hyperplasia. J Urol
1993;150:351-358.
112. Abrams P, D'Ancona C, Griffiths D, van Kerrebroeck P, Nishizawa O, Nitti V,
I l l
Keong Tatt F, Tubaro A, Wein A, Belal M. Lower urinary tract symptom:
etiology, patient assessment and predicting outcome from therapy. In:
McConnell J, Abrams P, Denis L, Khoury S, Roehrborn C, editors. Male Lower
Urinary Tract Dysfunction - Evaluation and Management. Paris: Health
Publications; 2005. p 69-142.
112
113. Gades NM, Jacobsen DJ, Girman CJ, Roberts RO, Lieber MM, Jacobsen SJ.
Prevalence of conditions potentially associated with lower urinary tract
symptoms in men. Br J Urol 2004;95(4):549-553.
114. Kortt MA, Bootman JL. The economics of benign prostatic hyperplasia
treatment: a literature review. Clin Ther 1996;18(6):1227-1241.
115. Jacobsen SJ, Girman CJ, Guess HA, Oesterling JE, Lieber MM. New
diagnostic and treatment guidelines for benign prostatic hyperplasia: potential
impact in the United States. Arch Int Med 1995;155(5):477-481.
116. Bautista OM, Kusek JW, Nyberg LM, McConnell JD, Bain RP, Miller G,
Crawford ED, Kaplan SA, Sihelnik SA, Brawer MK, Lepor H. Study design of
the Medical Therapy of Prostatic Symptoms (MTOPS) trial. Cont Clin Trials
2003;24(2):224.
117. St. Sauver JL, Sarma AV, Jacobson DJ, McGree ME, Lieber MM, Girman CJ,
Nehra A, Jacobsen SJ. Associations between C-Reactive Protein and Benign
Prostatic Hyperplasia/Lower Urinary Tract Symptom Outcomes in a
Population-based Cohort. Am J Epidemiol 2009;169(11):1281-1290.
118. Van Zee KJ, Kohno T, Fischer E, Rock CS, Moldawer LL, Lowry SF. Tumor
necrosis factor soluble receptors circulate during experimental and clinical
inflammation and can protect against excessive tumor necrosis factor a in vitro
and in vivo. Proc Natl Acad Sci USA 1992;89(11):4845-4849.
119. Englemann H, Aderka D, Rubinstein M, Rotman D, Wallach D. A tumor
necrosis factor-binding protein purified to homogeneity from human urine
protects cells from tumor necrosis factor toxicity. J Biol Chem
1989;264(20):11974-11980.
120. Sutcliffe S, Rohrmann S, Giovannucci E, Nelson KE, De Marzo AM, Isaacs
WB, Nelson WG, Platz EA. Viral infections and lower urinary tract symptoms
in the Third National Health and Nutrition Examination Survey. J Urol
2007;178(8):2181-2185.
121. Aderka D. The potential biological and clinical significance of the soluble
tumor necrosis factor receptors. Cytokine Growth Factor Rev 1996;7(3):231-
240.
122. Vandenabeele P, Declercq W, Beyaert R, Fiers W. Two tumour necrosis factor
113
receptors: structure and function. Trends Cell Biol 1995;5(10):392-399.
123. Holund B. Latent prostatic cancer in a consecutive autopsy series. Scandinavian
Journal of Urology and Nephrology 1980;14(l):19-35.
124. Breslow N, Chan CW, Dhom G, Drury AB, Franks LM, Gellei B, S LY,
Lundberg S, Sparke B, H SN, Tullinius HJ. Latent carcinoma of prostate at
autopsy in seven areas. Int J Cancer 1977;22:680-688.
125. Meigs JB, Barry MJ, Giovannucci E, Rimm EB, Stampfer MJ, Kawachi I. High
rates of prostate-specific antigen testing in men with evidence of benign
prostatic hyperplasia. Am J Med 1998;104:517-525.
126. Carroll P, Coley C, McLeod D, Schellhammer P, Sweat G, Wasson J, Zietman
A, Thompson I. Prostate-specific antigen best practice policy. Part I: Early
detection and diagnosis of prostate cancer. Urology 2001;57:217-224.
127. McNutt LA, Chuntao W, Xue X, Hafner JP. Estimating the relative risk in
cohort studies and clinical trials of common outcomes. Am J Epidemiol
2003;157(10):940-943.
128. Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence
ratios and differences. Am J Epidemiol 2005;162:199-200.
129. Gong Z, Kristal AR, Schenk JM, Tangen CM, Goodman PJ, Thompson IM.
Alcohol consumption, finasteride, and prostate cancer risk. Results from the
Prostate Cancer Prevention Trial. Cancer 2009;115:3661-3669.
130. Gong Z, Agalliu I, Lin DW, Stanford JL, Kristal AR. Cigarette smoking and
prostate cancer-specific mortality following diagnosis in middle-aged men.
Cancer Causes Control 2008;19:25-31.
131. Jacobsen SJ, Guess HA, Panser O, Girman CJ, Chute CG, Oesterling JE, Lieber
MM. A population-based study of health care-seeking behavior for treatment of
urinary symptoms. Arch Fam Med 1993;2:729-735.
132. Marks LS, Dorey FJ, Rhodes T, Shery ED, Rittenhouse H, Partin AW,
DeKernion JB. Serum prostate specific antigen levels after transurethral
resection of prostate: a longitudinal characterizationin men with benign
prostatic hyperplasia. J Urol 1996;156:1035-1039.
114
VITA
Jeannette Kisser (formerly Schenk) was born in Anaheim, California and permanently
resides in Seattle, Washington. She earned a Bachelor of Science degree in Molecular
Biology from the University of California at San Diego, and a Master of Science
degree in Clinical Nutrition from New York University. In 2010 she earned a Doctor
of Philosophy at the University of Washington in Epidemiology

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close