Urological Open Access Journal, 2012, 2, 20-27
Published Online February 2012 (http://www.SciRP.org/journal/uoaj)
The Effect of Obesity and Diabetes on Intermediate to
High Grade Prostate Cancer Patients Treated
with Radical Prostatectomy
Emma H. Ramahi1*, Katherine C. Ansley1*, Gregory P. Swanson1,2,3, Fei Du1,4, Joseph W. Basler1,3
1The University of Texas Health Science Center at San Antonio, San Antonio, Texas
2Department of Radiation Oncology, San Antonio, Texas
3Department of Urology, San Antonio, Texas
4Department of Epidemiology and Biostatistics, San Antonio, Texas
Received December 14, 2011; revised January 20, 2012; accepted January 30, 2012
ABSTRACT
Aims: The relationships between obesity, diabetes and prostate cancer are unclear. A retrospective study was performed
to determine the effects of body mass index (BMI) and diabetes on patients with intermediate to high grade prostate
cancer treated with radical prostatectomy. Methods: We reviewed 582 patients with Gleason score ≥ 7 non-metastatic
prostate cancer treated with radical prostatectomy. Patients were stratified by BMI. End points were biochemical failure
free survival (BFFS), overall survival (OS), and cancer specific survival (CSS). Results: Mean pre-treatment PSA de-
creased with increasing BMI (12.5, 7.6, 7.8 and 5.3 ng/mL with BMI < 25, 25 - 30, 30 - 35, and > 35, respectively; p <
0.001). There was no significant difference in BFFS, OS or CSS between diabetic and non-diabetic patients. After ad-
justing for potential confounders (age, Gleason score and pre-treatment PSA), patients with higher BMI experienced
biochemical failure more often with hazard ratios 1.87 (1.15, 3.04; p = 0.01), 2.23 (1.39, 3.56; p < 0.001), and 2.5 (1.22,
5.12; p = 0.01) for BMI 25 - 30, 30 - 35 and > 35, respectively. However, for overall mortality the adjusted hazard ratio
was 0.39 (0.18, 0.82; p = 0.01) for overweight patients (BMI 25 - 30) compared to patients with a BMI in the normal
range. Patients with a BMI of 30 - 35 and > 35 had increased rates of positive margins than those with a BMI of 25 - 30
or < 25 (41.4% and 45.5% versus 28.9% and 33.3%, respectively; p = 0.02). Patients with higher BMI than 25 had
higher biochemical recurrence rate (25 - 30 HR 2; 30 - 35 HR 1.97 and > 35 2.04) on multivariate analysis, margin
positivity alone was not a significant factor. Conclusions: Patients with increasing BMI tend to have a lower PSA at
diagnosis but are more likely to have biochemical failure after radical prostatectomy. In our cohort, this was not due to
the increased incidence of positive margins. Having diabetes had no effect.
Keywords: Obesity; Diabetes Mellitus; Prostatectomy; Prostatic Neoplasms; Therapy; Treatment Outcomes
1. Introduction
Prostate cancer and obesity are important causes of
morbidity and mortality in the United States. Prostate
cancer is the most commonly diagnosed non-cutaneous
cancer in American men, with approximately 220,000
cases identified each year [1]. It is estimated that prostate
cancer will represent 28% of new male cancer diagnoses
and will be the second leading cause of cancer death in
men in 2010 [1].
Approximately one third of American men are clini-
cally obese, and the prevalence of obesity has risen 6 to 7
percent each decade since the 1980s [2]. Obesity is a
well-known risk factor for a variety of chronic conditions
including hypertension, high cholesterol, stroke, heart
disease, and certain cancers such as colon, kidney, and
esophageal [3]. As obesity has grown more prevalent, the
incidence of type 2 diabetes mellitus has also increased,
and obesity increases the risk of death from diabetes up
to 9 times [4].
Faced with the increased incidence of prostate cancer
and obesity-related health issues, several investigators
have focused on how an increased body mass index
(BMI) and the presence of diabetes impact the incidence,
diagnosis, and response to treatment for prostate cancer.
Several studies show that obesity affects prostate cancer
mortality rather than incidence [5,6]. For example, in the
RTOG 85 - 31 trial of patients with locally advanced
prostate cancer undergoing radiation, it was found that 5-
year prostate cancer specific mortality for men with a
BMI < 25 was 6.5% compared with 13.1% and 12.2% in
*Ramahi EH and Ansley KC contributed equally to this work.

E. H. RAMAHI ET AL. 21
men with BMI ≥ 25 - 30 and BMI ≥ 30 respectively [7].
Therefore a greater baseline BMI was independently
associated with risk of death from prostate cancer. More
recent research, however, have not supported these find-
ings, so the association remains uncertain [8,9].
In addition, there is some evidence that shows type 2
diabetes is significantly linked with a decreased risk of
developing prostate cancer [10-12], but few studies have
evaluated the impact of diabetes on prostate cancer mor-
tality. One recent study evaluated 112 diabetic metformin
users and 98 diabetic non-metformin users treated with
radical prostatectomy and demonstrated that diabetics,
regardless of metformin use, faced a 55% increase in risk
of biochemical recurrence [13]; this was even though
there is some evidence that metformin may have some
antineoplastic effects [14].
Given the uncertainty regarding the effects of obesity,
diabetes and diabetes treatment on prostate cancer out-
come, we reviewed the outcome of our patients with
Gleason 7 - 10 cancer that were treated with radical pros-
tatectomy for these associations.
2. Materials and Methods
With institutional review board approval, we reviewed
patients in the South Texas Veteran’s Healthcare System
Tumor Registry diagnosed with prostate cancer between
January 1, 1998 and December 31, 2008. We entered
patients with Gleason scores greater than or equal to 7 on
biopsy or on radical prostatectomy pathology into a data-
base. 123 patients had a biopsy Gleason score < 7, but
were found to have Gleason 7 - 10 disease on radical
prostatectomy pathology. Statistical analysis was per-
formed both with and without the inclusion of these
patients, and their inclusion did not lead to significant
differences in outcome measure comparison.
Patient information including age at diagnosis, BMI at
the time of treatment initiation, race, presence or absence
of diabetes, the HbA1c value recorded closest to the time
of diagnosis, and choice of treatment for diabetes,
classified as either non-pharmacologic, metformin, other
oral medication or insulin were all obtained from the
medical record. Patients were coded to have diabetes
when the diagnosis was made and recorded in the medical
record by the patient’s primary physician prior to
initiation of treatment for prostate cancer. Also recorded
were the PSA immediately prior to diagnosis, information
from the original biopsy report including the Gleason
score as determined by the original attending pathologist,
and whether positive cores came from one or both sides
of the prostate. Information from the surgical pathology
report was also recorded including surgical Gleason
score as determined by the original attending pathologist,
the presence of disease in seminal vesicles (SV) or lymph
nodes (LN), the presence of extracapsular extension
(ECE), and the presence of positive margins. Patients
were excluded from analysis if they had metastatic
disease at the time of diagnosis as identified on bone
scan, computed tomography scan or magnetic resonance
imaging. Patients were also excluded from analysis if
their BMI was not recorded in the medical record or if
they were lost to follow up before post-treatment PSA
values could be obtained.
Biochemical failure was defined by the American
Urological Association Prostate Cancer Guideline Panel’s
definition of a PSA of 0.2 ng/mL or greater followed by
a confirmatory PSA value 0.2 ng/mL or greater [15].
Patients receiving a second modality as salvage treatment
were also considered to have failed on the date salvage
treatment was initiated, even if the above criteria had not
been met.
Patients were stratified into groups based on BMI into
the following categories: normal (BMI < 25), overweight
(BMI 25 - 29.9), obese (BMI 30 - 35) and morbidly
obese (BMI > 35). Additionally, patients were stratified
based on the presence or absence of the diagnosis of
diabetes at the time of treatment initiation, whether or not
their diabetes was controlled as evidence by HbA1c ≤ 7
or > 7, and by the treatment for diabetes: non-pharmaco-
logic, metformin, other oral agents, or insulin.
3. Statistical Methods
Continuously distributed data were summarized with the
sample size, mean and standard deviation (SD) and cate-
gorical data were described with counts and percentages.
Data were grouped by BMI category and diabetic status.
Groups were contrasted on continuously distributed out-
comes with Kruskal-Wallis tests. The significance of
associations between categorical outcomes and BMI ca-
tegory, diabetic status, and treatment were assessed with
Pearson’s chi-square test. BMI categories were con-
trasted with regard to biochemical free and overall sur-
vival with proportional hazards models and Kaplan
Meier survival curves and associated log rank tests. Haz-
ard ratios and their 95% confidence intervals are reported.
Hazard ratios were adjusted for potential confounders:
age, Gleason score, pre-treatment PSA, the presence of
diabetes and surgical pathology. All statistical testing
was 2-sided with a significance level of 5%. SAS Ver-
sion 9.2 for Windows (SAS Institute, Cary, North Caro-
lina) was used throughout.
4. Results
Mean follow up was 5.1 ± 2.8 years. A total of 582 pa-
tients with Gleason score ≥ 7 were analyzed by BMI and
diabetes category. Not surprisingly, there was a signifi-
cant trend of increasing incidence of diabetes with in-

E. H. RAMAHI ET AL.
22
Patients with the lowest (<25) BMI were found to have
significantly lower unadjusted 5-year OS (86.3% com-
pared to 93.6%, 94.6% and 93.2% for patients with BMI
of < 25, 25 - 29.9, 30 - 35 and > 35, respectively (p =
0.05) (Table 3). Even after controlling for age, Gleason
score, pre-treatment PSA and surgical pathology, moder-
ately overweight patients with a BMI of 25 - 30 and 30 -
35 had a lower overall mortality than those with a normal
BMI (<25) as evidenced by a hazard ratio of 0.38 (0.18,
0.79; p = 0.01) and 0.48 (0.23, 1.0; p = 0.05). Those with
a BMI > 35 were nonsignificantly higher (p = 0.92) with
a HR of 1.06 (0.37, 3.04) However, Kaplan-Meier esti-
mates did not show a significant difference in unadjusted
OS among the BMI groups.
creasing BMI. Specifically, for men with BMI > 35,
more than 50% were diabetic, while the incidence was
one third or less in those with BMI ≤ 35. There was a
significant decrease in PSA at diagnosis as BMI in-
creased; at the extremes, men with BMI < 25 had an av-
erage PSA of 12.5 ng/ml, while those with BMI > 35 had
an average PSA of 5.3 ng/ml. There was no significant
difference for Gleason score or the presence of bilateral
disease found on biopsy (Table 1). There was also no
significant difference between the BMI groups in terms
of RP surgical pathology features such as positive lymph
nodes, positive seminal vesicles and extracapsular exten-
sion. There were, however, significantly higher rates of
positive margins in patients with a BMI of 30 - 35 and >
35 when compared to patients with a BMI of < 25 and 25
- 30 (Table 2). 5. Discussion
BMI did not significantly affect unadjusted 5-year
BFFS or CSS (Table 3). Kaplan-Meier logistical regres-
sion analysis likewise showed no significant difference in
unadjusted BFFS (Figure 1). Neither the presence of
diabetes, the degree of control as evidenced by HbA1c at
diagnosis, nor the treatment regimen for diabetes had a
significant effect on 5 year BFFS, OS or CSS.
We were able to demonstrate an inverse relationship
between BMI and PSA at diagnosis, supporting the mul-
tiple groups who have observed that men with prostate
cancer and an increased BMI have a lower PSA [16-18]
than thinner men with similar appearing cancers. An arti-
ficially low PSA in an obese man potentially could delay
a prostate biopsy recommendation and increase risk for a
higher-grade cancer [19]. One recent hypothesis for the
low PSA in obese men faults hemodilution as a result of
increased plasma volume [20-22]. In our cohort, patients
are usually referred to urology for any elevation of the
PSA by a large number of providers, but there could be
some unknown selection factor that results in obese men
being referred earlier (with a lower PSA). In obese men
diagnosed with prostate cancer, Davies et al found that
they are less likely to receive surgical treatment [23].
There are several possible reasons why a physician and
their obese patient may decide against pursuing surgery
Patients using metformin had a 61.3% 5 yr BFFS,
while for the non users (not using metformin) it was 71%
(p = 0.09). There was no difference in CSS or OS.
After controlling for age, Gleason score, pre-treatment
PSA, margin, SV, and node positivity, patients with in-
creasing BMI above 25 were found to be at significantly
increasing risk of biochemical failure with hazard ratios
2 (1.21, 3.33; p = 0.007), 1.97 (1.21, 3.21; p = 0.006),
and 2.04 (0.98, 4.26; p = 0.06) for BMI 25 - 30, 30 - 35
and > 35, respectively (Table 4). Positive margins alone
did not explain this with no significant difference in haz-
ard rations for the different BMI groups. (Table 4).
Figure 1. Unadjusted biochemical failure free survival by body mass index.

E. H. RAMAHI ET AL. 23
Table 1. Pretreatment characteristics for patients undergoing radical prostatectomy for Gleason 7 - 10 prostate cancer.
BMI: <25 25 - 30 30 - 35 >35 Total p-value1
Age 0.151
N 117 218 203 44 582
Mean (SD) 62.4 (6.9) 63.6 (6.9) 62.6 (6.7) 60.6 (6.7) 62.8 (6.8)
Median (Q1, Q3) 62.2 (57.8, 66.7) 62.9 (58.9, 67.9) 61.9 (58.6, 67.4) 61.6 (56.1, 65) 62.6 (58.4, 67.3)
Min, Max 44.04, 79 46.56, 88.1 43.73, 78.6 36.7, 71.8 36.7, 88.1
Race, n (%) 0.462
Unkown 10 (8.5) 37 (17) 32 (15.8) 6 (13.6) 85 (14.6)
White 96 (82.1) 154 (70.6) 150 (73.9) 34 (77.3) 434 (74.6)
Black 10 (8.5) 24 (11) 16 (7.9) 4 (9.1) 54 (9.3)
Other 1 (0.9) 3 (1.4) 5 (2.5) 0 (0) 9 (1.5)
Total 117 218 203 44 582
PSA at diagnosis <0.0011
N 111 214 194 40 559
Mean (SD) 12.5 (23) 7.6 (8.8) 7.8 (7.2) 5.3 (2.2) 8.5 (12.5)
Median (Q1, Q3) 6.7 (5.1, 11.3) 5.5 (4.2, 8.3) 5.4 (4.2, 8.4) 4.9 (4.2, 6.6) 5.6 (4.3, 8.6)
Min, Max 0.74, 221.9 1.01, 99.6 0.52, 48.1 1.17, 12.7 0.52, 221.9
Gleason Score, n (%) 0.552
7 57 (60) 119 (70) 108 (67.9) 21 (60) 305 (66.4)
8 21 (22.1) 24 (14.1) 30 (18.9) 8 (22.9) 83 (18.1)
9 - 10 17 (17.9) 27 (15.9) 21 (13.2) 6 (17.1) 71 (15.5)
Total 95 170 159 35 459
Bilateral disease, n (%) 0.972
Unilateral 61 (53) 117 (54.2) 103 (52.3) 24 (55.8) 305 (53.4)
Bilateral 54 (47) 99 (45.8) 94 (47.7) 19 (44.2) 266 (46.6)
Total 115 216 197 43 571
Diabetes, n (%) <0.0012
No 104 (88.9) 170 (78) 136 (67) 21 (47.7) 431 (74.1)
Yes 13 (11.1) 48 (22) 67 (33) 23 (52.3) 151 (25.9)
Total 117 218 203 44 582
HbA1c 0.581
N 12 47 66 23 148
Mean (SD) 6.7 (1.5) 6.7 (1.3) 6.9 (1.6) 6.9 (0.9) 6.8 (1.4)
Median (Q1, Q3) 6.1 (5.8, 7) 6.5 (5.8, 7.2) 6.5 (5.9, 7.5) 6.6 (6.3, 7.5) 6.5 (5.9, 7.4)
Min, Max 5.6, 10 4.9, 10.2 5, 12.4 5.6, 8.9 4.9, 12.4
Diabetes Treatment, n (%) <0.0012
Metformin Use 3 (2.6) 21 (9.6) 36 (17.7) 15 (34.1) 75 (12.9)
Insulin 2 (1.7) 8 (3.7) 4 (2) 1 (2.3) 15 (2.6)
Other 4 (3.4) 7 (3.2) 12 (5.9) 3 (6.8) 26 (4.5)
No treatment 108 (92.3) 182 (83.5) 151 (74.4) 25 (56.8) 466 (80.1)
Total 117 218 203 44 582
1Kruskal-wallis Test; 2Pearson’s Chi-square Test.

E. H. RAMAHI ET AL.
24
Table 2. Surgical pathology by body mass index in patients undergoing radical prostatectomy for Gleason 7 - 10 prostate
cancer.
BMI
Surgical Pathology <25 25 - 30 30 - 35 >35 Total p-value1
+ LN, N (%) 0.752
No 112 (95.7) 212 (97.2) 198 (97.5) 42 (95.5) 564 (96.9)
Yes 5 (4.3) 6 (2.8) 5 (2.5) 2 (4.5) 18 (3.1)
Total 117 218 203 44 582
+ SV, N (%) 0.392
No 104 (88.9) 193 (88.5) 176 (86.7) 35 (79.5) 508 (87.3)
Yes 13 (11.1) 25 (11.5) 27 (13.3) 9 (20.5) 74 (12.7)
Total 117 218 203 44 582
+ ECE, N (%) 0.352
No 87 (74.4) 176 (80.7) 155 (76.4) 31 (70.5) 449 (77.1)
Yes 30 (25.6) 42 (19.3) 48 (23.6) 13 (29.5) 133 (22.9)
Total 117 218 203 44 582
+ Margins, N (%) 0.022
No 78 (66.7) 155 (71.1) 119 (58.6) 24 (54.5) 376 (64.6)
Yes 39 (33.3) 63 (28.9) 84 (41.4) 20 (45.5) 206 (35.4)
Total 117 218 203 44 582
BMI = body mass index, LN = lymph nodes, SV = seminal vesicles, ECE = extracapsular extension; 1Kruskal-wallis Test; 2Pearson’s Chi-Square Test.
Table 3. 5-year survival outcomes by body mass index, presence, control and treatment of diabetes.
5-year BFFS % p-value1 5-year OS % P-value1 5-year CSS % p-value1
BMI
<25 75.2% 0.46 86.3% 0.05 100% 0.54
25 - 29.9 69.3% 93.6% 98.6%
30 - 35 67% 94.6% 99%
>35 72.7% 93.2% 97.7%
Diabetes
No 71.4% 0.13 91.9% 0.81 98.6% 0.78
Yes 65.6% 93.4% 99.3%
HbA1C
≤7 65.4% 0.74 95.2% 0.15 100% 0.12
>7 68.2% 88.6% 97.7%
Diabetes Treatment
Metformin Use 61.3% 0.38 96% 0.08 98.7% 0.76
Insulin 80% 80% 100%
Other 65.4% 88.5% 100%
1Kruskal-wallis Test.
as the primary therapeutic modality. Primarily, they tend
to have less favorable outcomes than do thinner men, and
there are several reasons possible reasons for this. Jaya-
chandran et al. postulate that these inferior outcomes are
due to the increased technical difficulty of the surgery in
heavier patients [24]. This is supported by our findings of

E. H. RAMAHI ET AL. 25
Table 4. Multivariate analysis for biochemical failure risk
by body mass index.
Control for Age, Gleason Score,
Pre-treatment PSA, Mar+, SV+
and LN+
Control for Mar+
BMI HR (95% CI) P HR (95% CI)P
<25 1.0 (reference) - 1.0 (reference)-
25 - 30 2 (1.21, 3.33) 0.007 1.18 (0.78, 1.79)0.43
30 - 35 1.97 (1.21, 3.21) 0.006 1.23 (0.81, 1.85)0.33
>35 2.04 (0.98, 4.26) 0.06 1.14 (0.61, 2.14)0.67
Age is continuous variable.
increased positive margins with increased BMI despite
no significant difference in Gleason score, the rate of po-
sitive lymph nodes, seminal vesicles or extracapsular
extension. These data support the finding that with larger
patients, there is an increase in positive margins. Although
we saw an increased risk of biochemical failure with
increasing BMI, this was not explained by the finding of
positive margins (Table 4). In fact, when corrected for
known prognostic factors, obese patients still had a
significantly higher failure rate. (Table 4).
While our data suggest these increasing rates of bio-
chemical failure do not translate into decreased overall or
cancer specific survival with our length of follow up, at
the minimum these patients will be subjected more often
to the morbidity of salvage treatments such as radiation
therapy, hormone therapy, or chemotherapy. Unlike the
RTOG 85 - 31 trial, we were unable to show worse CSS
in patients with increased BMI. This is in agreement with
the findings of two recent studies [8,9]. These studies
have either shown no difference [8] or a trend toward
decreasing OS with increasing BMI [9]. Kane et al ana-
lyzed patients in the Cancer of the Prostate Strategic
Urologic Research Endeavor (CaPSURE) registry and
found that overweight and obese patients with prostate
cancer tended to have more medical comorbidities such
as diabetes and hypertension. Kane et al. also hypothe-
sized that overweight and obese men may have an in-
creased diagnosis of low risk prostate cancer due to more
frequent doctor’s visits for their other co-morbid condi-
tions [18]. Our data showed increasing rates of diabetes
as BMI increased, so close medical follow up for this and
other potential chronic medical condition may explain
our finding of decreased mortality in the overweight BMI
group (BMI 25 - 30) even after controlling for potential
confounders such as age, Gleason score, pre-treatment
PSA and the presence of diabetes. We also sought to ex-
plore the effect of diabetes and its treatment on 5-year
OS, BFFS and CSS. Interestingly, some studies suggest
diabetics may have a decreased risk of developing pros-
tate cancer, which is thought to be due to multiple hor-
monal factors [25]. This raises the question as to whether
those with cancer might fare better than non-diabetic
patients. Unfortunately, Patel showed that diabetes, re-
gardless of metformin use, resulted in a 55% increase in
risk of biochemical failure [13]. However, we were un-
able to demonstrate any significant difference in outcome
between diabetics and non-diabetics. In addition, even
though Metformin is thought to exert both anti-tumor and
anti-proliferative effects [26] and has been associated
with decreased total cancer risk in epidemiologic studies,
our patients did not appear to benefit from its use. If any-
thing, in our patients, metformin users have a higher
biochemical recurrence rate. Study limitations include
those inherent to any retrospective analysis. Selection of
treatment was at the discretion of the original treating
physician. Also, our single HbA1c level at diagnosis is
only a snapshot of diabetes control, so a detailed analysis
of disease outcome based on that parameter is not possi-
ble. Additionally, a mean follow up of 5.1 years is still
too short to make definitive statements regarding the
ultimate development of metastatic disease and prostate
cancer mortality. Further follow up will make this more
clear.
6. Conclusions
Previous studies have raised the concern that increased
BMI and diabetes adversely affect prostate cancer out-
comes. As the average BMI and the incidence of diabetes
continue to increase, these issues will affect more and
more men. Our data suggest that although overweight
and obese men tend to be younger and with a lower PSA
at diagnosis, they are more likely to have co-morbid
diagnoses such as diabetes. Most importantly, obese men
are more likely to have positive surgical margins, which
may contribute to the increased rate of biochemical
recurrence after radical prostatectomy. Our data show no
clear relationship between BMI and prostate cancer
specific survival. There is uncertainty in the current
literature as to the exact effect of obesity and diabetes on
prostate cancer. Large scale prospective randomized
controlled trials would be required to resolve these issues.
Ultimately, the molecular mechanisms linking obesity,
diabetes, and prostate cancer are multifactorial. In the
meantime, as the morbidity and mortality associated with
obesity and diabetes are well established, we should
continue to encourage weight loss, increased physical
activity, and glycemic control in our patients with pro-
state cancer.
7. Acknowledgements
The authors would like to thank Joel Michalek, PhD of
the University of Texas Health Science Center at San
Antonio Department of Epidemiology and Biostatistics
for his assistance in editing this manuscript.

E. H. RAMAHI ET AL.
26
REFERENCES
[1] A. Jemal, R. Siegel, J. Xu and E. Ward, “Cancer Statistics,
2010,” Cancer Journal for Clinicians, Vol. 60, No. 5,
2010, pp. 277-300. doi:10.3322/caac.20073
[2] K. M. Flegal, M. D. Carroll, C. L. Ogden and L. R. Curtin,
“Prevalence and Trends in Obesity among US Adults,
1999-2008,” Journal of the American Medical Associa-
tion, Vol. 303, No. 3, 2010, pp. 235-241.
doi:10.1001/jama.2009.2014
[3] E. E. Calle, C. Rodriguez, K. Walker-Thurmond and M. J.
Thun, “Overweight, Obesity, and Mortality from Cancer
in a Prospectively Studied Cohort of U.S. Adults,” The
New England Journal of Medicine, Vol. 348, No. 17,
2003, pp. 1625-1638. doi:10.1056/NEJMoa021423
[4] R. G. Rogers, R. A. Hummer and P. M. Krueger, “The
Effect of Obesity on Overall, Circulatory Disease- and
Diabetes-Specific Mortality,” Journal of Biosocial Sci-
ence, Vol. 35, No. 1, 2003, pp. 107-129.
doi:10.1017/S002193200300107X
[5] A. R. Kristal and Z. Gong, “Obesity and Prostate Cancer
Mortality,” Future Oncology, Vol. 3, No. 5, 2007, pp.
557-567. doi:10.2217/14796694.3.5.557
[6] P. Dimitropoulou, R. M. Martin, E. L. Turner, et al., “As-
sociation of Obesity with Prostate Cancer: A Case-Con-
trol Study within the Population-Based PSA Testing
Phase of the Protect Study,” British Journal of Cancer,
Vol. 104, No. 5, 2011, pp. 875-881.
doi:10.1038/sj.bjc.6606066
[7] J. A. Efstathiou, K. Bae, W. U. Shipley, et al., “Obesity
and Mortality in Men with Locally Advanced Prostate
Cancer: Analysis of RTOG 85-31,” Cancer, Vol. 110, No.
12, 2007, pp. 2691-2699. doi:10.1002/cncr.23093
[8] B. J. Davies, M. C. Smaldone, N. Sadetsky, et al., “The
Impact of Obesity on Overall and Cancer Specific Sur-
vival in Men with Prostate Cancer,” Journal of Urology,
Vol. 182, No. 1, 2009, pp. 112-117.
doi:10.1016/j.juro.2009.02.118
[9] J. Pfitzenmaier, M. Pritsch, A. Haferkamp, et al., “Is the
Body Mass Index a Predictor of Adverse Outcome in
Prostate Cancer after Radical Prostatectomy in a Mid-
European Study Population?” British Journal of Urology
International, Vol. 103, No. 7, 2009, pp. 877-882.
doi:10.1111/j.1464-410X.2008.08149.x
[10] S. Bonovas, V. Peponis and K. Filioussi, “Diabetes Mel-
litus as a Risk Factor for Primary Open-Angle Glaucoma:
A Meta-Analysis,” Diabetic Medicine, Vol. 21, No. 6,
2004, pp. 609-614.
doi:10.1111/j.1464-5491.2004.01173.x
[11] J. S. Kasper and E. Giovannucci, “A Meta-Analysis of
Diabetes Mellitus and the Risk of Prostate Cancer,” Can-
cer Epidemiology, Biomarkers & Prevention, Vol. 15, No.
11, 2006, pp. 2056-2062.
doi:10.1158/1055-9965.EPI-06-0410
[12] E. A. Atchison, G. Gridley, J. D. Carreon, et al., “Risk of
Cancer in a Large Cohort of U.S. Veterans with Diabe-
tes,” International Journal of Cancer, Vol. 128, No. 3,
2011, pp. 635-643. doi:10.1002/ijc.25362
[13] T. Patel, G. Hruby, K. Badani, et al., “Clinical Outcomes
after Radical Prostatectomy in Diabetic Patients Treated
with Metformin,” Urology, Vol. 76, No. 5, 2010, pp.
1240-1244. doi:10.1016/j.urology.2010.03.059
[14] J. M. Evans, L. A. Donnelly, A. M. Emslie-Smith, et al.,
“Metformin and Reduced Risk of Cancer in Diabetic Pa-
tients,” British Medical Journal, Vol. 330, No. 7503,
2005, pp. 1304-1305. doi:10.1136/bmj.38415.708634.F7
[15] M. S. Cookson, G. Aus, A. L. Burnett, et al., “Variation
in the Definition of Biochemical Recurrence in Patients
Treated for Localized Prostate Cancer: The American
Urological Association Prostate Guidelines for Localized
Prostate Cancer Update Panel Report and Recommenda-
tions for a Standard in the Reporting of Surgical Out-
comes,” Journal of Urology, Vol. 177, No. 2, 2007, pp.
540-545. doi:10.1016/j.juro.2006.10.097
[16] L. L. Banez, L. Sun, B. J. Trock, et al., “Body Mass In-
dex and Prostate Specific Antigen as Predictors of Ad-
verse Pathology and Biochemical Recurrence after Pro-
statectomy,” Journal of Urology, Vol. 182, No. 2, 2009,
pp. 491-496. doi:10.1016/j.juro.2009.04.007
[17] J. Baillargeon, B. H. Pollock, A. R. Kristal, et al., “The
Association of Body Mass Index and Prostate-Specific
Antigen in a Population-Based Study,” Cancer, Vol. 103,
No. 5, 2005, pp. 1092-1095. doi:10.1002/cncr.20856
[18] C. J. Kane, W. W. Bassett, N. Sadetsky, et al., “Obesity
and Prostate Cancer Clinical Risk Factors at Presentation:
Data from CaPSURE,” Journal of Urology, Vol. 173, No.
3, 2005, pp. 732-736.
doi:10.1097/01.ju.0000152408.25738.23
[19] N. Parekh, Y. Lin, R. S. Dipaola, et al., “Obesity and
Prostate Cancer Detection: Insights from Three National
Surveys,” American Journal of Medicine, Vol. 123, No. 9,
2010, pp. 829-835. doi:10.1016/j.amjmed.2010.05.011
[20] L. L. Banez, R. J. Hamilton, A. W. Partin, et al., “Obe-
sity-Related Plasma Hemodilution and PSA Concentra-
tion among Men with Prostate Cancer,” Journal of the
American Medical Association, Vol. 298, No. 19, 2007,
pp. 2275-2280. doi:10.1001/jama.298.19.2275
[21] R. L. Grubb, A. Black, G. Izmirlian, et al., “Serum Pro-
state-Specific Antigen Hemodilution among Obese Men
Undergoing Screening in the Prostate, Lung, Colorectal,
and Ovarian Cancer Screening Trial,” Cancer Epidemi-
ology, Biomarkers & Prevention, Vol. 18, No. 3, 2009, pp.
748-751. doi:10.1158/1055-9965.EPI-08-0938
[22] S. Culp and M. Porter, “The Effect of Obesity and
Lower Serum Prostate-Specific Antigen Levels on Pros-
tate-Cancer Screening Results in American Men,” British
Journal of Urology International, Vol. 104, No. 10, 2009,
pp. 1457-1461. doi:10.1111/j.1464-410X.2009.08646.x
[23] B. J. Davies, T. J. Walsh, P. L. Ross, et al., “Effect of
BMI on Primary Treatment of Prostate Cancer,” Urology,
Vol. 72, No. 2, 2008, pp. 406-411.
doi:10.1016/j.urology.2007.11.032
[24] J. Jayachandran, W. J. Aronson, M. K. Terris, et al.,
“Obesity and Positive Surgical Margins by Anatomic
Location after Radical Prostatectomy: Results from the
Shared Equal Access Regional Cancer Hospital Data-
base,” British Journal of Urology International, Vol. 102,
No. 8, 2008, pp. 964-968.

E. H. RAMAHI ET AL.
27
doi:10.1111/j.1464-410X.2008.07881.x
[25] J. S. Kasper, Y. Liu, M. N. Pollak, et al., “Hormonal Pro-
file of Diabetic Men and the Potential Link to Prostate
Cancer,” Cancer Causes Control, Vol. 19, No. 7, 2008,
pp. 703-710. doi:10.1007/s10552-008-9133-x
[26] I. Ben Sahra, K. Laurent, S. Giuliano, et al., “Targeting
Cancer Cell Metabolism: The Combination of Metformin
and 2-Deoxyglucose Induces p53-Dependent Apoptosis
in Prostate Cancer Cells,” Cancer Research, Vol. 70, No.
6, 2010, pp. 2465-2475.
doi:10.1158/0008-5472.CAN-09-2782
[27] A. Decensi, M. Puntoni, P. Goodwin, et al., “Metformin
and Cancer Risk in Diabetic Patients: A Systematic Re-
view and Meta-Analysis,” Cancer Prevention Research,
Vol. 3, No. 11, 2010, pp. 1451-1461.
doi:10.1158/1940-6207.CAPR-10-0157
[28] C. J. Currie, C. D. Poole and E. A. Gale, “The Influence
of Glucose-Lowering Therapies on Cancer Risk in Type 2
Diabetes,” Diabetologia, Vol. 52, No. 9, 2009, pp. 1766-
1777. doi:10.1007/s00125-009-1440-6