Author's Accepted Manuscript

Increased risk of cancer in infertile men: Analysis of US claims data

Michael L. Eisenberg , Shufeng Li , James D. Brooks , Mark R. Cullen , Laurence

C. Baker

PII: S0022-5347(14)04926-X
DOI: 10.1016/j.juro.2014.11.080
Reference: JURO 12014

To appear in: The Journal of Urology

Accepted Date: 10 November 2014

Please cite this article as: Eisenberg ML, Li S, Brooks JD, Cullen MR, Baker LC, Increased risk of
cancer in infertile men: Analysis of US claims data, The Journal of Urology® (2014), doi: 10.1016/
j.juro.2014.11.080.

DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a
service to our subscribers we are providing this early version of the article. The paper will be copy edited
and typeset, and proof will be reviewed before it is published in its final form. Please note that during the
production process errors may be discovered which could affect the content, and all legal disclaimers
that apply to The Journal pertain.

Embargo Policy

All article content is under embargo until uncorrected proof of the article becomes available
online.

We will provide journalists and editors with full-text copies of the articles in question prior to the embargo
date so that stories can be adequately researched and written. The standard embargo time is
12:01 AM ET on that date. Questions regarding embargo should be directed to jumedia@elsevier.com.
 ACCEPTED MANUSCRIPT

Increased risk of cancer in infertile men: Analysis of US claims data

Michael L. Eisenberg, MD,1,2 Shufeng Li, MS,3 James D. Brooks, PhD,1 Mark R.

PT
Cullen, MD4, Laurence C. Baker, PhD5

RI
1 Department of Urology; Stanford University School of Medicine; Stanford,

SC
California

2 Department of Obstetrics/Gynecology; Stanford University School of Medicine;

U
Stanford, California
AN
3 Departments of Urology and Dermatology; Stanford University School of

Medicine; Stanford, California

M

4 Department of Internal Medicine; Stanford University School of Medicine,

D

Stanford, California
TE

5 Department of Health Research and Policy; Stanford University School of

Medicine, Stanford, California

C EP
AC

Word Count: 2,580

Key Words: male infertility, fertility, cancer

Corresponding Author:

Michael L. Eisenberg
Department of Urology
 ACCEPTED MANUSCRIPT

Stanford University School of Medicine

300 Pasteur Drive
Stanford, CA 94305-5118
Phone: 650-723-5700
Fax: 650-498-5346
Email: eisenberg@stanford.edu

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Abstract

Purpose

Aberrations in reproductive fitness may be a harbinger of medical diseases in

PT
men. Data suggest a higher risk of testicular cancer in infertile men. However,

RI
the relationship between infertility and other cancers remains uncertain.

SC
Materials and Methods

U
We analyzed subjects contained within the Truven Health MarketScan claims
AN
database from 2001 to 2009. Infertile men were identified through diagnosis and

treatment codes. Comparison groups of vasectomized men and a “control”

M

cohort of men not infertile and not vasectomized were created. The incidence of
D

cancer was compared to national US estimates. Infertile men were also

TE

compared to vasectomized men and the control cohort utilizing a Cox regression

model.
EP

Results
C
AC

In all, 76,083 infertile men were identified with an average age of 35.1 years. A

total of 112,655 vasectomized men and 760,830 control men were assembled.

Compared to age-adjusted national averages, infertile, vasectomized and control

men in the study cohorts all had higher rates of all cancers and many individual

cancers. In time to event analysis, infertile men had a higher risk for cancer than
 ACCEPTED MANUSCRIPT

to vasectomized or control men. Infertile men had a higher risk of testis cancer,

non Hodgkins lymphoma, and all cancers compared to both vasectomized and

control men.

PT
Conclusion

RI
Consistent with prior reports, we identified an elevated risk of testicular cancer

SC
in infertile men. The current data also suggest that infertile men are at a mildly

elevated risk of all cancers in the years following infertility evaluation. Future

U
research should focus on confirming these associations and elucidating pathways
AN
between infertility and cancer.
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Introduction

Aberrations in reproductive fitness may be a harbinger of medical diseases in

PT
men. Indeed, recent studies from Europe and the U.S. suggest that male

infertility is associated with an increased risk of cancer in the years following an

RI
infertility evaluation. To date, most studies support an increased risk of testis

SC
cancer in subfertile men. Using paternity as a surrogate for fertility, a Danish

case control study demonstrated a higher risk of testis cancer in men with fewer

U
offspring.1 This group then later confirmed these findings in a cohort of infertile
AN
men by showing that men with impaired semen quality had a higher risk of

subsequently developing testis cancer (SIR 1.6, 95% CI 1.3-1.9).2 Walsh and
M

colleagues examined a cohort of infertile couples in California and found that

D

men with male factor infertility (as defined by a treating provider) had an
TE

increased risk for testis cancer (SIR 2.8, 95% CI 1.5-4.8).3

While infertile men are thought to be at higher risk for testis cancer, the risks
EP

of other malignancies are less certain. Studying the California infertility cohort,
C

Walsh and colleagues also showed men with male factor infertility were at
AC

higher risk for high grade prostate cancer, though overall prostate cancer risk

was unaltered.4 In contrast, a Swedish case control study found that infertile

men were at a lower risk of prostate cancer.5 In 1992, a US group identified a

trend associating male factor infertility with breast cancer in a questionnaire-

based case control study.6 Reasoning that azoospermia, the most severe form of
 ACCEPTED MANUSCRIPT

male factor infertility, would have the highest risk of adverse health outcomes,

our group previously examined cancer incidence in men with azoospermia and

found that azoospermic men had an increased risk of all cancers (SIR 2.9, 95% CI

PT
1.4-5.4).7

Investigators have hypothesized that genetic, hormonal, environmental,

RI
lifestyle, or in utero factors could explain a link between a man’s fertility and

SC
cancer risk.8, 9 As approximately 10% of the male human genome is involved in

reproduction, it is not unlikely that other health conditions may be linked to

U
impaired fertility.10, 11 However, the relationship between fertility and cancer is
AN
difficult to track given the low incidence of cancer in reproductive aged men and

the difficulty of following reproductive outcomes without centralized registries.

M

Since many private insurance organizations do cover infertility testing, in this

D

study we sought to utilize a large database of health insurance claims to examine

TE

outcomes in men with male infertility. Given the biologic plausibility of this

association, the conflicting data in the literature, and the relevance to millions of
EP

men impacted by infertility; we sought to determine if a diagnosis of male factor

C

infertility was associated with subsequent cancer risk.

AC

Materials and Methods

Patients
 ACCEPTED MANUSCRIPT

We analyzed subjects contained within the Truven Health Marketscan™

Commercial Claims and Encounters database. This database provides

information from adjudicated and paid insurance claims filed for the care of

PT
privately-insured individuals with employment-based insurance through a

participating employer. Marketscan provides claims data on 77 million covered

RI
lives since 1996. This study used data from 2001 through 2009. The number of

SC
individuals represented in the database varies over time; the more recent years of

the data contain more than 30 million covered lives.

U
We focus on a cohort of likely infertile men, identified by the presence on
AN
outpatient claims of an infertility diagnosis code (International Statistical

Classification of Diseases, 9th edition (ICD9) 606.x, V26.21) or by the presence on

M

any claim with a procedure code (Current Procedural Terminology (CPT)) for
D

fertility testing or semen analysis/semen preparation (89300, 89310, 89320, 89321,

TE

89322, 89325, 89329, 89330, 89331). We recorded the first date of a relevant

diagnosis or procedure code as the index date. Given the variable infertility
EP

coding and reimbursement practices in the US, we attempted to be as broad

C

with our definition as possible.

AC

As a cancer diagnosis and treatment can lead to additional cancers, men with

any claim with a diagnosis code for cancer prior to the index date or within one

year after the index date were excluded. Subjects were required to be enrolled in

a plan covered by the database for at least one year preceding the index date, and
 ACCEPTED MANUSCRIPT

for greater than one year following the index date. Subjects were required to be

between the ages of 18 and 50 on the index date.

A comparison group of men age 18-50 with claims containing a procedure

PT
code for vasectomy (CPT 55250 or 55450) was assembled, as this group should

include few or no infertile men.12 Men in this group were assigned an index date

RI
as the earliest date of a claim with a vasectomy procedure code, and were

SC
required to have enrollment in a plan covered by the database for at least one

year prior to and one year after the index date.

U
In addition, a “control” group of men not in either of the two previously
AN
described cohorts was assembled. For this cohort, we selected 10 men for each

man in the infertile cohort, matched on age and follow-up time.

M

For each man in the cohorts, the number of outpatient visits after the
D

index date was ascertained based on the presence of claims with CPT codes
TE

indicating new and follow up office visits, consultations, or preventative

medicine encounters. Medical comorbidities were determined based on ICD9

EP

codes on any claim and included hypertension (401- 405), obesity (278.0),

smoking (305.1, V1582), and diabetes (250 - 250.93).

C
AC

Outcome ascertainment

Cancer diagnoses were identified using diagnosis codes on inpatient and

outpatient claims. Diagnosis codes identifying cancer were aligned to SEER

 ACCEPTED MANUSCRIPT

definitions. We identified men with claims diagnoses indicating the presence of

any invasive cancer (ICD9 140-209 excluding skin squamous cell, skin basal cell

and non invasive cancers), and identified men with codes indicating the presence

PT
of specific cancers: upper respiratory (140.x-149.x, 160.x, 161.x), stomach (151.x),

colorectal (153.x, 154.0, 154.1, 154.8), liver and gallbladder (155.x and 156.x),

RI
pancreas (157.x), lung (162.x), melanoma (172.x), breast (175.x), prostate (185.x),

SC
testis (186.x), urinary bladder (188.x), kidney (189.0, 189.1), brain and nervous

system (191.x, 192.x), thyroid (193.x), non Hodgkins Lymphoma (NHL; 200.x,

U
202.x), Hodgkins Lymphoma (HL, 201.x), Leukemia (204.x, 205.x, 206.x, 207.x,
AN
208.x), and esophageal (150.x).
M

Statistical analysis
D
TE

Men accrued at-risk time beginning one year after their index dates, until

cancer diagnosis or December 31 of the final year of enrollment in a health plan

EP

included in the MarketScan database. The first year was excluded since a cancer
C

diagnosis within this time period was an exclusion criteria.

AC

We compared cancer incidence rates in our cohorts with cancer incidence

in the general U.S. population using Standardized Incidence Rates (SIRs), for all

cancers and for each specific type of cancer. We calculated the expected number

of cancers for our cohorts by multiplying the number of years at risk by the 5-
 ACCEPTED MANUSCRIPT

year age strata death rates from SEER,13 and derived the SIRs by dividing the

observed number of cancers by the expected number of cancers.

We then compared the risk of cancer in infertile men to risk in the vasectomy

PT
cohort and the control cohort using a Cox proportional hazard regression model

that adjusted for age, year of evaluation, comorbidity, and follow-up time. All p

RI
values were two sided with p<0.05 considered statistically significant. Analyses

SC
were performed using SAS (version 9.3, SAS Institute, Inc, Cary, NC).

U
Results
AN
In all, 47,385 men had a diagnosis of male factor infertility, and an additional
M

28,698 received fertility or semen testing. Combining these created a total of

D

76,083 in our infertile cohort. The average age of the infertile men was 35.08 (SD
TE

5.89), and there was a total of follow up time of 277,703 person years. A total of

112,655 men were identified as having a vasectomy. In addition, 760,830 men

EP

were included in the control cohort of men who were neither infertile nor
C

vasectomized (Table 1).

AC

Infertile men and vasectomized men had more comorbidities compared to the

control group. In addition, infertile men and vasectomized men had a higher

number of average visits per year during follow up compared the control

group, which may explain the discrepancies in comorbidities. Less than 0.1
 ACCEPTED MANUSCRIPT

visits per year on average were coded as being related to male infertility.

Moreover, while 16% of infertile men and 15% of vasectomized men had no

outpatient visits during follow up, 20% of control men never received medical

PT
care. Characteristics of the cohort are given in table 1.

RI
During follow up, 600 infertile men, 919 vasectomized men, and 3,986 control

SC
men developed cancer. Compared to age adjusted SEER estimates, infertile,

vasectomized and control men all had higher rates of all cancers and many

U
individual cancers (Table 2).
AN
We next compared cancer risks in our three cohorts. After adjustment, infertile
M

men had the highest risk for cancer (Table 3). Compared to controls, infertile
D

men had a 49% higher risk of all cancers during follow up and nearly a two fold
TE

higher risk of testis cancer. In addition, the risk of melanoma, prostate cancer,

bladder cancer, thyroid cancer, HL, NHL, and leukemia were all higher in
EP

infertile men. Vasectomized men had a 22% higher risk of all cancers compared
C

to control men with no difference in the risk of testis cancer. The risks of prostate
AC

cancer and melanoma were higher for vasectomized men. Lastly, we compared

the risk of developing cancer between infertile and vasectomized men. Infertile

men had higher rates of any cancer (HR 1.19, 1.07-1.32), testis cancer (1.50, 1.01-

2.22), and NHL (1.71, 1.25-2.32). Separate analysis of the infertile groups

identified by diagnosis codes or by procedure codes did not alter the overall
 ACCEPTED MANUSCRIPT

conclusions for any of the analyses. Adjustment for co-morbidities did not

materially impact the results.

PT
Discussion

RI
We identified a higher risk of cancer in privately insured men age 18-50 who

SC
were likely to be infertile, in the years following an infertility diagnosis or

evaluation. While incidence rates for many cancers were elevated in infertile

U
men as well as in our two control groups, infertile men had the highest rates
AN
among the three groups. Specifically, the risk of all cancers, testicular cancer,

and non Hodgkin’s Leukemia was higher in infertile men compared to both
M

control groups.
D

Consistent with prior reports, we identified an elevated risk of testicular

TE

cancer in infertile men. Importantly, the elevated risk estimate, as assessed by

comparison to SEER data or internal cohort controls ranged from 50-115% which
EP

is consistent with the current literature.1-3, 14 As such, analysis of the infertile men
C

within the MarketScan cohort corroborates existing data relating the established
AC

relationship between male infertility to an elevated risk of testicular cancer.

The association between male infertility and other cancers is less clearly

established in previous literature. Danish investigators found that testis cancer

was the only cancer associated with semen quality after linking semen data of

32,000 men with the national cancer registry.2 While a study of infertile
 ACCEPTED MANUSCRIPT

California couples noted a higher incidence of prostate cancer in men with male

factor infertility, a Swedish study did not find such an association.4, 5 In fact,

several US and European studies have shown that childless men have a lower

PT
risk of prostate cancer compared to fathers.15-17

The current report noted a higher rate of several cancers compared to age

RI
adjusted estimated from SEER. However, both control groups, also showed

SC
higher cancer rates compared to those expected from the SEER data suggesting

that the population represented in the MarketScan database may be different

U
from that represented in the SEER registries. It is important to note that SEER
AN
represents data from several distinct geographical populations.18 While

MarketScan does collect data from the entire United States, the specific
M

location of individual patients was not available and may impact results.
D

Given the geographic variations in cancer rates, the inability to adjust for
TE

subject location represents a limitation in being able to adequately compare to

SEER population estimates. For that reason, we also compared our infertile
EP

subjects to internal controls also derived from the MarketScan data (i.e. men who

had a vasectomy or men who had neither a vasectomy or infertility). While each
C
AC

population is distinct, we reasoned that comparisons of these groups would

provide greatest insight. While care utilization was similar between infertile and

vasectomized men, the control group had significantly lower rates of annual

clinician visits which can also impact cancer through screening and diagnostic

factors.19-21 In addition, the infertile and vasectomized men were older and had
 ACCEPTED MANUSCRIPT

higher rates of comorbidities compared to the control group both of which may

lead to higher rates of cancer incidence.22-24 Interestingly, while the

vasectomized men were older and generally had higher rates of comorbidities, it

PT
was the infertile men had higher rates of cancer. Owing to these limitations, we

have focused on those categories whose rates differ across our three comparator

RI
groups (i.e. SEER, control, and vasectomy). Following this approach, infertile

SC
men showed higher rates of all cancers, testis cancer, and non Hodgkin’s

lymphoma. The higher rate of prostate cancer in vasectomized men has been

U
previously explored and has been plausibly attributed to screening bias.25, 26
AN
Our study, like other observational studies, is limited by unmeasured

confounders between groups, such as differences in the number of doctor

M

visits and screening tests, that prohibit assignment of cause. This limitation is
D

highlighted by the absence of a plausible biological connection between

TE

vasectomy and prostate cancer.

The link between male infertility and cancer remains uncertain although
EP

several convincing explanations have been proposed. As up to 10% of the male

genome is involved in reproduction, it is likely that other nonprocreative

C

processes may also be affected by aberrations in fertility.10, 11 In addition,

AC

investigators have hypothesized that fetal exposures may leading to adverse

reproductive and somatic health in adulthood.9, 27 While infertility may be a

marker of diminished fitness, which may accelerate the development of impaired

health in the future, it may also occur as a consequence of current health, clinical or
 ACCEPTED MANUSCRIPT

subclinical. As such, pre-clinical chronic diseases may diminish male fertility

and lead to impaired health later in life. Indeed, obesity and smoking can both

reduce fertility and increase cancer incidence.22-24, 28 Moreover, cancers evolve

PT
over many years before the become clinically manifest, and it is also possible that

some occult carcinomas could impair fertility years before they are detected,

RI
possibly through secretions of hormones, growth factors, cytokines or other

SC
factors.

Several additional limitations warrant mention. While the number of men

U
and total years at risk was high, the average follow up for each individual and
AN
final age was limited. Moreover, men could not be tracked after leaving the

database. Given the low rate of cancer in reproductive aged men, our outcome
M

rate was modest. As with many analyses that rely on administrative data,
D

granular information on each man regarding race, infertility diagnosis, and

TE

lifestyle factors was unavailable. In addition, MarketScan represents only

commercially insured individuals and thus may not represent all US men.29
EP

Finally, differences in care utilization, as evidenced by rates of annual visits,

may impact cancer diagnoses because of observational biases.30

C
AC

Nevertheless, the current study demonstrates the potential of claims based

data in understanding the relationship of male reproductive health to future

somatic health. Despite the relative short follow-up of the cohort, the large

populations available still allowed the confirmation of established relationships

(i.e. male infertility and testis cancer) and identified two novel associations (i.e.
 ACCEPTED MANUSCRIPT

male infertility and all cancers or NHL). The current data suggest that infertile

men are at a mildly elevated risk of all cancers in the years immediately

following infertility evaluation or treatment. Future research should focus on

PT
confirmation of these associations and elucidation of the pathways between

infertility and cancer.

RI
SC
References:

U
AN
1. Moller, H., Skakkebaek, N. E.: Risk of testicular cancer in subfertile men:

case-control study. BMJ, 318: 559, 1999

M

2. Jacobsen, R., Bostofte, E., Engholm, G. et al.: Risk of testicular cancer in

D

men with abnormal semen characteristics: cohort study. BMJ, 321: 789,
TE

2000

3. Walsh, T. J., Croughan, M. S., Schembri, M. et al.: Increased risk of

EP

testicular germ cell cancer among infertile men. Arch Intern Med, 169: 351,
C

2009
AC

4. Walsh, T. J., Schembri, M., Turek, P. J. et al.: Increased risk of high-grade

prostate cancer among infertile men. Cancer, 116: 2140, 2010

5. Ruhayel, Y., Giwercman, A., Ulmert, D. et al.: Male infertility and prostate

cancer risk: a nested case-control study. Cancer causes & control : CCC,

21: 1635, 2010

 ACCEPTED MANUSCRIPT

6. Thomas, D. B., Jimenez, L. M., McTiernan, A. et al.: Breast cancer in men:

risk factors with hormonal implications. Am J Epidemiol, 135: 734, 1992

7. Eisenberg, M. L., Betts, P., Herder, D. et al.: Increased cancer risk and

PT
azoospermia. Fertil Steril, 100: 681, 2013

8. Hotaling, J. M., Walsh, T. J.: Male infertility: a risk factor for testicular

RI
cancer. Nature reviews. Urology, 6: 550, 2009

SC
9. Skakkebaek, N. E., Rajpert-De Meyts, E., Main, K. M.: Testicular

dysgenesis syndrome: an increasingly common developmental disorder

U
with environmental aspects. Hum Reprod, 16: 972, 2001
AN
10. Matzuk, M. M., Lamb, D. J.: The biology of infertility: research advances

and clinical challenges. Nat Med, 14: 1197, 2008

M

11. Hotaling, J., Carrell, D. T.: Clinical genetic testing for male factor
D

infertility: current applications and future directions. Andrology, 2: 339,

TE

2014

12. Eisenberg, M. L., Henderson, J. T., Amory, J. K. et al.: Racial differences in

EP

vasectomy utilization in the United States: data from the national survey
C

of family growth. Urology, 74: 1020, 2009

AC

13. Surveillance Research Program, N.: The Surveillance, Epidemiology, and

End Results (SEER) Program of the National Cancer Institute, 2014

14. Doria-Rose, V. P., Biggs, M. L., Weiss, N. S.: Subfertility and the risk of

testicular germ cell tumors (United States). Cancer causes & control : CCC,

16: 651, 2005

 ACCEPTED MANUSCRIPT

15. Eisenberg, M. L., Park, Y., Brinton, L. A. et al.: Fatherhood and incident

prostate cancer in a prospective US cohort. Int J Epidemiol, 40: 480, 2011

16. Giwercman, A., Richiardi, L., Kaijser, M. et al.: Reduced risk of prostate

PT
cancer in men who are childless as compared to those who have fathered a

child: a population based case-control study. Int J Cancer, 115: 994, 2005

RI
17. Jorgensen, K. T., Pedersen, B. V., Johansen, C. et al.: Fatherhood status and

SC
prostate cancer risk. Cancer, 112: 919, 2008

18. Warren, J. L., Klabunde, C. N., Schrag, D. et al.: Overview of the SEER-

U
Medicare data: content, research applications, and generalizability to the
AN
United States elderly population. Medical care, 40: IV, 2002

19. Andriole, G. L., Crawford, E. D., Grubb, R. L., 3rd et al.: Mortality results
M

from a randomized prostate-cancer screening trial. The New England

D

journal of medicine, 360: 1310, 2009

TE

20. Schroder, F. H., Hugosson, J., Roobol, M. J. et al.: Screening and prostate-

cancer mortality in a randomized European study. The New England

EP

journal of medicine, 360: 1320, 2009

C

21. Garg, V., Raisch, D. W., Selig, J. P. et al.: Health disparities in clinical
AC

practice patterns for prostate cancer screening by geographic regions in

the United States: a multilevel modeling analysis. Prostate cancer and

prostatic diseases, 16: 193, 2013

 ACCEPTED MANUSCRIPT

22. Adams, K. F., Leitzmann, M. F., Albanes, D. et al.: Body size and renal cell

cancer incidence in a large US cohort study. Am J Epidemiol, 168: 268,

2008

PT
23. Koebnick, C., Michaud, D., Moore, S. C. et al.: Body mass index, physical

activity, and bladder cancer in a large prospective study. Cancer

RI
epidemiology, biomarkers & prevention : a publication of the American

SC
Association for Cancer Research, cosponsored by the American Society of

Preventive Oncology, 17: 1214, 2008

U
24. Freedman, N. D., Leitzmann, M. F., Hollenbeck, A. R. et al.: Cigarette
AN
smoking and subsequent risk of lung cancer in men and women: analysis

of a prospective cohort study. The lancet oncology, 9: 649, 2008

M

25. Howards, S. S., Peterson, H. B.: Vasectomy and prostate cancer. Chance,
D

bias, or a causal relationship? JAMA, 269: 913, 1993

TE

26. Siddiqui, M. M., Wilson, K. M., Epstein, M. M. et al.: Vasectomy and Risk

of Aggressive Prostate Cancer: A 24-Year Follow-Up Study. Journal of

EP

clinical oncology : official journal of the American Society of Clinical

C

Oncology, 2014
AC

27. Barker, D. J.: Fetal origins of coronary heart disease. Bmj, 311: 171, 1995

28. Sermondade, N., Faure, C., Fezeu, L. et al.: BMI in relation to sperm count:

an updated systematic review and collaborative meta-analysis. Human

reproduction update, 19: 221, 2013

 ACCEPTED MANUSCRIPT

29. Hotaling, J. M., Davenport, M. T., Eisenberg, M. L. et al.: Men who seek

infertility care may not represent the general U.S. population: data from

the National Survey of Family Growth. Urology, 79: 123, 2012

PT
30. Wennberg, J. E., Staiger, D. O., Sharp, S. M. et al.: Observational intensity

bias associated with illness adjustment: cross sectional analysis of

RI
insurance claims. Bmj, 346: f549, 2013

U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table Legends

Table 1. Characteristics of cohort

PT
Table 2. Age adjusted standardized incidence rates of cancers compared to SEER

RI
data for all infertile (diagnosis and treatment codes), infertile (ICD9 coded male

SC
factor infertility), vasectomy, and control (neither infertility nor vasectomy).

U
Table 3. Cox regression models comparing infertile, vasectomy, and control
AN
groups for all cancers. Hazard ratio (95% confidence interval) listed. All models

adjusted for age, year of evaluation, comorbidity, and follow up time.

M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 1. Characteristics of cohort

Infertile

PT
Infertile
Control (All diagnoses and Vasectomy
(Male factor)
treatment)

RI
n 760,830 76,083 47,385 112,655
Age (years), mean

SC
(SD) 18-19 1110 (0.15) 111 (0.15) 53 (0.11) 14 (0.01)
20-29 133310(17.52) 13331(17.52) 8390 (17.71) 10324 (9.16)
30-39 452740(59.51) 45274(59.51) 28550 (60.25) 64372 (57.14)

U
40-50 173670(22.83) 17367(22.83) 10392 (21.93) 37945 (33.68)

AN
Follow up time
(years), mean (SD) 1-2 194110(25.51) 19411(25.51) 11999 (25.32) 25131 (22.31)

M
2-3 235110 (30.90) 23511 (30.90) 13895 (29.32) 33764 (29.97)
3-4 152770 (20.08) 15277 (20.08) 9183 (19.38) 22974 (20.39)

D
4+ 178840 (23.51) 17884 (23.51) 12308 (25.97) 30786 (27.33)

TE
Hypertension 124099 (16.31) 13678 (17.98) 8755(18.48) 22058(19.58)
Hyperlipidemia 53858 (7.08) 14165 (18.62) 9025(19.05) 23039(20.45)
EP
Obesity 4837 (0.64) 1340 (1.76) 888(1.87) 1689(1.50)
Smoking 7593 (1.00) 1255 (1.65) 812(1.71) 2605(2.31)
Diabetes 16266 (2.14) 1907 (2.51) 1304(2.75) 2461(2.18)
C
AC

Year of evaluation, 2001 40156 (5.28) 2724(3.58) 2003(4.23) 4034(3.58)

n (%) 2002 81763 (10.75) 5151 (6.77) 3573(7.54) 8372(7.34)
2003 84535 (11.11) 7404 (9.73) 4929(10.40) 11779(10.46)
2004 117223 (15.41) 11793 (15.50) 7930(16.74) 16892(14.99)
 ACCEPTED MANUSCRIPT

2005 91683 (12.05) 11529 (15.15) 7599(16.04) 17874(15.87)

2006 205809 (27.05) 19188 (25.22) 11194(23.62) 26713(23.71)
2007 139661 (18.36) 18294 (24.04) 10157(21.44) 26991(23.96)

PT
Outpatient visits Avg visits per person yr 1.83(2.35) 2.55 (3.03) 2.60 (3.06) 2.48 (2.76)

RI
Men with 0 visits 152379 (20.03) 12497 (16.43) 7487 (15.80) 16683 (14.81)
Men with 0-1 visit/yr 164548 (21.63) 8150(10.71) 5016 (10.59) 13007 (11.55)

SC
Men with 1-2 visits/yr 175922 (23.12) 17245(22.67) 10756 (22.70) 26515 (23.54)
Men with 2+ visits/yr 267981 (35.24) 38191 (50.20) 24126 (50.91) 56450 (50.11)

U
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 2. Age adjusted standardized incidence rates of cancers compared to SEER data for all infertile (diagnosis and
treatment codes), infertile (ICD9 coded male factor infertility), vasectomy, and control (neither infertility nor vasectomy).

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

Table 3. Cox regression models comparing infertile, vasectomy, and control groups for all cancers. Hazard ratio (95%
confidence interval) listed. All models adjusted for age, year of evaluation, comorbidity, and follow up time.

PT
Cancers Infertile vs Control Infertile vs Vasectomy Vasectomy vs Control
all malignant 1.49 (1.37 - 1.63) 1.19 (1.07 - 1.32) 1.22 (1.13 - 1.31)

RI
Upper aerodigestive tract 0.99 (0.65 - 1.51) 0.80 (0.49 - 1.29) 1.23 (0.92 - 1.65)
Stomach 0.76 (0.27 - 2.12) 0.47 (0.15 - 1.46) 1.39 (0.77 - 2.52)

SC
Colon and Rectum 1.32 (0.97 - 1.80) 1.12 (0.77 - 1.62) 1.11 (0.86 - 1.43)
Liver and gallbladder 1.35 (0.72 - 2.54) 1.79 (0.75 - 4.28) 0.63 (0.33 - 1.23)
Pancreas 1.05 (0.48 - 2.31) 1.26 (0.46 - 3.42) 0.79 (0.40 - 1.56)

U
Lung 0.71 (0.40 - 1.25) 0.54 (0.29 - 1.03) 1.10 (0.78 - 1.57)

AN
Melanoma 1.37 (1.06 - 1.77) 0.78 (0.58 - 1.04) 1.72 (1.43 - 2.07)
Breast 1.09 (0.33 - 3.62) 0.74 (0.18 - 3.10) 1.31 (0.53 - 3.21)
Prostate 1.78 (1.41 - 2.25) 1.22 (0.93 - 1.59) 1.44 (1.19 - 1.74)

M
Testis 1.99 (1.47 - 2.70) 1.50 (1.01 - 2.22) 1.27 (0.94 - 1.73)
Urinary bladder 2.29 (1.49 - 3.50) 1.49 (0.89 - 2.49) 1.42 (0.96 - 2.11)

D
Kidney 0.93 (0.58 - 1.48) 0.88 (0.51 - 1.51) 1.07 (0.76 - 1.52)
Brain + Other nervous

TE
1.34 (0.97 - 1.86)
system 1.16 (0.75 - 1.79) 0.80 (0.48 - 1.33)
Thyroid 1.52 (1.01 - 2.30) 1.23 (0.75 - 2.03) 1.20 (0.83 - 1.72)
EP
NHL 1.76 (1.39 - 2.23) 1.71 (1.25 - 2.32) 0.99 (0.78 - 1.26)
HL 1.67 (1.04 - 2.67) 1.15 (0.64 - 2.06) 1.30 (0.84 - 1.99)
Leukemia 1.82 (1.29 - 2.59) 1.54 (0.98 - 2.40) 1.11 (0.79 - 1.57)
C

Esophagus 1.03 (0.41 - 2.60) 1.11 (0.34 - 3.63) 1.05 (0.52 - 2.10)
AC
 ACCEPTED MANUSCRIPT

Key of Definitions for Abbreviations:

CI – confidence interval

PT
CPT – Current Procedural Terminology

ICD9 - International Statistical Classification of Diseases, 9th edition

RI
SIR – standardized incidence rate

U SC
AN
M
D
TE
C EP
AC