CIGARETTE SMOKING EXPOSURE AND ITS EFFECT ON CONVENTIONAL SEMEN
PARAMETERS

By

Yasaman Osati Back

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

Epidemiology - Master of Science

2013

ABSTRACT

CIGARETTE SMOKING EXPOSURE AND ITS EFFECT ON CONVENTIONAL SEMEN
PARAMETERS

By

Yasaman Osati Back

Cigarette smoke has been associated with many cancers and other health conditions and there is
concern about the possible negative effects of smoking on semen parameters and male
reproduction. This study reports on the effect of cigarette smoking on semen parameters from
data collected as part of the Fish and Infertility Study (FINS), a cross-sectional NIH study
undertaken to evaluate the environmental factors on measures of male infertility. 603 men
between the ages of 18 and 60 were recruited from couples presenting at two infertility clinics in
Michigan. Participants filled out a detailed questionnaire on lifestyle factors and provided
semen, blood, and urine specimens. Data from the FINS study was analyzed to assess smoking
exposure and other lifestyle variables that may negatively affect sperm parameters. We found
significant and increased odds of low normal sperm morphology between participants who were
“Ever Smokers” versus “Never Smokers” (OR=1.61, p=0.032, CI= 1.043, 2.496). No significant
associations were found between cigarette smoke exposure and total sperm count, sperm
concentration, semen volume, or motility. When total sperm count and sperm concentration
were assessed, exercise was shown to have a significant negative effect on both semen
parameters, while work stress seemed to function as a protective factor. We also detected a
protective association between the wearing of boxers and sperm motility.

To my incredible family and loving husband whose encouragement and continuous support made
this accomplishment possible.

iii

TABLE OF CONTENTS

LIST OF TABLES

v

LIST OF FIGURES

viii

INTRODUCTION

1

BACKGROUND
Epidemiologic Studies
Cross-Sectional Studies
Case-Control Studies
Other Studies
Cotinine Studies
Summary
Commentary

3
5
5
6
7
8
9
10

METHODS
Objectives
Study Design
Study Population
Data Collection
Analysis of Semen Specimens
Data Analysis

16
16
16
17
17
18
18

RESULTS

20

DISCUSSION

24

APPENDICES
Tables
Figures

31
32
63

BIBLIOGRAPHY

70

iv

LIST OF TABLES

Table 1. Classification of low or abnormal semen parameters.

32

Table 2. Body Mass Index (BMI) categories and calculation.

33

Table 3. Variables used in analyses, by unit and type.

34

Table 4. Outcomes and exposure used in multivariable logistic regression models.

35

Table 5. Demographic characteristics of participants.

36

Table 6. Socio-economic characteristics of participants.

37

Table 7. Smoking characteristics of participants.

38

Table 8. Semen parameters of participants.

39

Table 9. Semen parameters of participants classified by “Normal” or “Abnormal”.

40

Table 10. Univariable analysis depicting mean semen parameters and cigarette smoke
exposure.

41

Table 11. Univariable analysis depicting mean semen parameters and subject
demographics.

42

Table 12. Univariable analysis depicting mean semen parameters and caffeine
consumption.

43

v

Table 13. Univariable analysis depicting mean semen parameters and alcohol consumption.
44
Table 14. Univariable analysis depicting mean semen parameters and other lifestyle
factors.

45

Table 15. Logistic regression model for Low Total Sperm Count, using “Smoke Now” as
the main variable (N=145).
46

Table 16. Logistic regression model for Low Total Sperm Count, using “Ever Smoked” as
the main variable (N=448).
47

Table 17. Logistic regression model for Low Total Sperm Count, using “Monthly Hours of
Passive Smoking” as the main variable (N=445).
48

Table 18. Logistic regression model for Low Total Sperm Count, using “Monthly Hours of
Passive Smoking” and “Smoke Now” as main variables (N=143).
49

Table 19. Logistic regression model for Low Sperm Concentration, using “Smoke Now” as
the main variable (N=141).
50

Table 20. Logistic regression model for Low Sperm Concentration, using “Ever Smoked”
as the main variable (N=425).
51

Table 21. Logistic regression model for Low Sperm Concentration, using “Monthly Hours
of Passive Smoking” as the main variable (N= 422).
52

Table 22. Logistic regression model for Low Semen Volume, using “Smoke Now” as the
main variable (N=216).
53

Table 23. Logistic regression model for Low Semen Volume, using “Ever Smoked” as the
main variable (N=535).
54

Table 24. Logistic regression model for Low Semen Volume, using “Monthly Hours of
Passive Smoking” as the main variable (N=531).

vi

55

Table 25. Logistic regression model for Low Semen Volume, using “Monthly Hours of
Passive Smoking” and “Smoke Now” as the main variables (N=213).

56

Table 26. Logistic regression model for Low Sperm Motility, using “Smoke Now” as the
main variable (N=150).
57

Table 27. Logistic regression model for Low Sperm Motility, using “Ever Smoked” as the
main variable (N=461).
58

Table 28. Logistic regression model for Low Sperm Motility, using “Monthly Hours of
Passive Smoking” as the main variable (N=458).

59

Table 29. Logistic regression model for Low Normal Sperm Morphology, using “Smoke
Now” as the main variable (N=212).
60

Table 30. Logistic regression model for Low Normal Sperm Morphology, using “Ever
Smoked” as the main variable (N= 530).

61

Table 31. Logistic regression model for Low Normal Sperm Morphology, using “Monthly
Hours of Passive Smoking” as the main variable (N=526).
62

vii

LIST OF FIGURES

Figure 1. Age of Study Subjects.

63

Figure 2. Body Mass Index (BMI) of Study Subjects.

64

6

Figure 3. Total sperm count of study subjects (10 ).

65
6

Figure 4. Semen concentration of study subjects (10 /mL).

66

Figure 5. Ejaculate volume of study subjects (mL).

67

Figure 6. Percent normal sperm motility of study subjects.

68

Figure 7. Percent normal sperm morphology of study subjects.

69

viii

INTRODUCTION
1

A substantial proportion of men of reproductive age worldwide smoke. Cigarette
smoking varies across regions, a staggering 23.4% of men in the United States take part
in smoking. Cigarette smoke has been associated with many cancers including bladder,
oral cavity, kidney, stomach, pancreas, and it has been causally linked to lung cancer.

2

Smoking has also been correlated with an increased risk of coronary heart disease and
stroke.

3,4,5

A number of elements have caused concern about the possible negative

effects of smoking on semen parameters and male reproduction. The considerable
prevalence of male smokers, the many adverse health conditions caused by smoking and
the fact that cigarette smoke contains more than 30 chemical agents known to be
mutagens, aneugens, or carcinogens add to the significance of this issue.

1

Alarmingly, semen quality has been declining in industrialized counties in the
past half century.

6, 7, 8, 9

A 2010 Spanish cross-sectional study found that sperm count

and sperm concentration may have declined in Spanish men over the last decade.

10

Another cross-sectional study found that in the general Danish population, only one in
four men had optimal semen quality and 25% of men will most likely face a prolonged
waiting time to pregnancy if they intend to father a child in the future and another 15%
are at risk of the need of fertility treatment.

11

Moreover, a French retrospective study

conducted with data from IVF clinics throughout 1989-2005 found a 32.3% decrease in
sperm concentrations, with projects that indicated that concentration for a 35 year old
man went from an average of 73.6 million/ml in 1989 to 49.9 million/ml in 2005.
1

12

Fertility rates have drastically declined in most European countries to below
replacement level of the population.

13, 14

Furthermore, diminished sperm quality has

been linked to decreased fecundity, which may in turn affect fertility rates.

15

Various

studies on the effects of cigarette smoking on semen parameters have been implemented,
but the scientific community has not yet reached a clear consensus. There are several
reasons why a consensus on the topic has not yet been reached, many of which appertain
to limitations in study designs. Small sample size, use of infertile study population, and
use of questionnaires to obtain exposure status are a few of the drawbacks in previous
studies.
Nonetheless, cigarette smoking is a voluntary activity that is overwhelmingly
detrimental to one’s health. Given the declining fertility rates in the general male
population, if cigarette smoking does have unfavorable effects on male reproduction and
fertility, it is important that public health measures are implemented to promote smoking
cessation with regards to improving male reproduction, especially for males with existing
infertility issues.

2

BACKGROUND
Cigarette smoking has increased worldwide throughout the past few decades.

16

Considering the fact that cigarette smoke is a carcinogen and a somatic cell mutagen,
there is plausible concern that paternal smoking may cause adverse reproductive
outcomes.

17

Reproductive capacity is dependent on multiple physiological and genetic
functions, and there is a multitude of ways in which cigarette smoking could have an
1

detrimental effect on male reproduction. A number of mechanisms could describe the
effects of compounds found in cigarette smoke on male reproduction: harmful effects of
18

cadmium and other heavy metals , mutagenic effects caused by aromatic
19

20

hydrocarbons , decreased accessibility to hemoglobin due to carbon monoxide ,
21

accrual of radioactive particles in the testes , and pernicious effects of nicotine.

22

Essentially, cigarette smoke could have a possible cytotoxic effect on sperm or a
reduction in sperm number and sperm functionality. Likewise, smoke from cigarettes
may cause sub-fertility by altering hormone levels. Testosterone levels may be reduced,
elevated or unchanged, and estradiol levels are largely found to be elevated in smokers.

23

Smoking may also damage the testes and other parts of the reproductive tract, leading to
impaired spermatogenesis.

1

Semen quality depends on multiple measures and is a good representation of
sperm production and function, and even fertility to some extent. The usual measures of
semen quality include total sperm count (the total number sperm per ejaculate), sperm
3

concentration (the number of sperm/ml of ejaculate), sperm volume (total volume of
ejaculate), sperm morphology (the evaluation of sperm shape and size; frequently
reported as the percentage of abnormally/normally shaped spermatozoa), and sperm
motility (the analysis of sperm movement, commonly quantified as the percentage of
motile sperm). Many of these measurements have been associated with reduced fertility.
Namely, a low total sperm count, a low percentage of sperm motility and a high
percentage of abnormally shaped sperm.

24

Various research studies have been conducted concerning the effect of cigarette
smoke on semen parameters. However, a clear consensus has not yet been reached.
Several studies have demonstrated that cigarette smoke has a negative impact on
conventional semen parameters. In contrast, a number of studies have concluded that an
association between cigarette smoking and its effect on semen variables does not exist.
For the purpose of this thesis, studies that were most current (no older than 1990) and
relevant were reviewed. Google Scholar was used in the search for papers using a variety
of amalgamations of the phrase “the effects of cigarette smoking” and following words or
terms: “sperm”, “semen parameters”, “sperm quality”, “sperm count”, “sperm
concentration”, “semen volume”, “sperm motility”, and “sperm morphology”.
Sometimes the word “semen” was used instead of “sperm”. Finally, more specific
searches were also executed in order to identify studies on the effects of caffeine, alcohol,
stress, and exercise on semen parameters.

4

Epidemiologic Studies
For this review, we found three cross-sectional studies, two case-control studies,
one study that was not categorized, and a systematic review on smoking and male
reproduction.
Cross-Sectional Studies
Trummer et al. implemented a cross-sectional study to compare semen parameters
and hormone concentrations of infertile smokers and infertile non- and ex-smokers.

25

1,104 men with infertility for ≥1 year were separated into three groups: 517 nonsmokers,
109 ex-smokers, and 478 smokers. Ex-smokers were defined as men who stopped
smoking ≥ 6 months prior to examination for infertility. Smokers were defined as men
who had smoked cigarettes for >6 months and were still smoking. In this paper
“infertility” was undefined. Cigarette smoking exposures were assessed via questionnaire
and semen analyses were performed following World Health Organization (WHO)
guidelines (Table 1). An increase in testosterone levels was observed in smoking
subjects. All smoking subjects were also urged, via free verbal counseling and mailed
anti-smoking information, to cease smoking in order to improve fertility prior to a second
semen analysis. However, significantly more non-smokers than smokers returned for the
second semen analysis, which led Trummer to conclude that only a few idiopathic
infertile smokers were able to quit smoking and that smoking did not affect conventional
semen parameters.
A cross-sectional study by Kunzle et al. revisited the effects of cigarette smoking
on main sperm variables by using a large cohort of men from an infertility clinic.

26

The

study population consisted of 655 smokers and 1,131 non-smokers. The criteria defining
5

“smokers” was strict (> 1 cigarette /day). Smoking and health histories were obtained
using questionnaires and semen analyses were performed by established methods. The
study exhibited an association between cigarette smoking and a significant decrease in
normal forms of sperm, sperm density (-15.3%), total sperm count (-17.5%), and total
number of motile sperm (-16.6%). It was concluded that cigarette smoking was
associated with reduced semen quality and that those who have borderline semen quality
who wish to have children could benefit from smoking cessation.
A cross-sectional study by Sepaniak et al. used a cohort of men from an infertility
clinic and the study population consisted of 57 non-smokers and 51 smokers.

27

As with

many studies, smoking status was self-reported and obtained by questionnaire. In this
study, however, a CO test was also used to confirm participant’s cigarette smoke
exposure. Sepaniak et al. did not find a significant difference in conventional parameters
between smokers and non-smokers.

27

They did, however, find a significant difference in

the DNA fragmentation rate between smokers and non-smokers (p< 0.01; 32% versus
25.9%, respectively). Sepaniak concluded that damaged sperm DNA could lessen the
chances of a successful pregnancy by damaging embryonic development and
implantation, or by increasing the rate of spontaneous abortion. Ultimately, smoking
cessation was recommended for anyone considering biological children in the future.
Case-Control Studies
Two case-control studies found that cigarette smoking adversely affected semen
functions and fertility. The first was a study set in the department of obstetric and
gynecology of a tertiary general hospital.

28

The study population was comprised of 218

6

infertile men (cases), and 240 fertile men (controls) whose wives were pregnant at the
time of the study. Smoking exposure status and health histories were collected through a
questionnaire obtained by a trained interviewer. Semen collections were performed at the
subject’s own home and brought into the hospital within one hour of collection. Semen
analyses were completed according to WHO guidelines. The study found that semen
parameters of all cases were significantly poorer than that of the controls and that the risk
of infertility is associated with smoking (crude OR 2.82, 95% CI 1.93-4.13; adjusted OR
2.96; 95% CI 1.98-4.42). The second case-control study took place in Shandong
Province, China.

29

The cases consisted of 110 non-smokers and 191 smokers, and the

controls were comprised of 61 fertile non-smokers who had one or more children. The
smokers were divided into subgroups according to the amount and the duration of
smoking. Smoking status was obtained by direct interview, and semen parameters were
examined using WHO guidelines. The results revealed that semen volume, density,
viability, and forward progression were much lower in the medium, heavy and long-term
28

smokers, respectively, than in non-smokers (P< 0.01). Similarly to Chia et al. , the
29

researchers of this case-control study

concluded that cigarette smoking affected semen

quality in the population of men in the study.
Other Studies
Sofikitis et al. examined the effects of smoking on testicular function, semen
quality and sperm fertilizing capacity.

30

In this study, 77 men undergoing left testicular

biopsy for surgical correction of a left inguinal hernia were divided into groups. Group 1
was composed of 49 smokers, age 21-38, who smoked more than 20 cigarettes per day
7

for longer than 3 years. Group 2 was comprised of 28 men, age 18-44, who had never
smoked. Finally, group 3 consisted of 9 men from Group 1 who stopped smoking 8-12
months postoperatively. Smoking exposure was obtained pre- and 6 months
postoperatively through a questionnaire, independently confirmed by the subject’s wives
or partners, and re-affirmed through a nicotine/cotinine/trans-3-hydroxycotinine
assessment. Several semen samples were collected postoperatively and a standard semen
analysis was performed. The study revealed that testosterone levels in the left testicular
vein, left testicular androgen-binding protein secretion rate (in vitro), sperm motility,
percentage of morphologically normal spermatozoa, sperm morphometric parameters and
outcome of sperm function tests were significantly lower (p< 0.05) in smokers than in
nonsmokers. Since testosterone and androgen-binding protein have important roles in
normal spermatogenesis and the epididymal sperm maturation process, respectively,
Sofikitis et al. postulated that reduced levels of the aforementioned factors may lead to a
decrease in sperm fertilizing capacity. The study also demonstrated that sperm motility
and morphology significantly improved in the 9 smokers who ceased smoking (Group 3).
These findings suggested a causal relationship between smoking and a decrease in sperm
motility and function.
A 1996 large-scale systematic review examined the effects of smoking and male
reproduction. The review found that cigarette smoking was associated with modest
reduction in sperm quality including sperm concentration, motility, and morphology.
Associations between male smoking and sperm concentration and motility were stronger
among healthy men (volunteers and sperm donors) than men from infertility clinic
populations. However, despite modest reductions in sperm quality, the review did not

8

show a reduction in male fertility in association with paternal smoking. It concluded that
although smokers as a group may not experience reduced fertility, men with questionable
sperm quality who wish to have children may benefit from quitting smoking, since
several studies indicated a potential for improved semen quality after smoking cessation.

1

Cotinine Studies
Various studies have focused on cotinine levels in relation to semen parameters
and most found a significant association between cotinine concentrations and semen
parameters. A cross-sectional study of 35 smoking students and 30 non-smoking
students in a university in Taiwan concluded that cotinine may decrease male fertility by
inhibiting density, reducing total progressively motile sperm count, and increasing the
percentage of abnormal sperm.

31

Similarly, a case-control study of 107 fertile and 103

sub-fertile men found a small but statistically significant correlation between cotinine
concentration in seminal plasma and the percentage of abnormal sperm morphology, but
not for other semen parameters.

32

A prospective study of 23 patients with normal

seminal parameters, 11 smokers, and 12 non-smokers concluded that cotinine levels were
correlated highly with the number of cigarettes smoked per day.

33

Yet another study

focused on the effects of cotinine on sperm fertilizing capacity in vitro and found that
cotinine concentrations of 400 or 800 ng/ml exerted detrimental effects on sperm,
motility, membrane function, and the ability to undergo capacitation.

34

Thus, using an

unbiased measure of smoking, adverse effects of smoking on sperm were still detected.

9

Summary
Two out of three cross-sectional studies did not find an association between
cigarette smoking and conventional sperm parameters.

25, 27

However, two of the studies

did recommend smoking cessation for men considering having children in the
future.

26,27

Both of the case-control studies reviewed found a significant inverse

association between cigarette smoking exposure and sperm quality.

28, 29

A study by

Sofikitis et al. suggested a causal relationship between smoking and a decrease in sperm
motility and function.

30

Finally, Vine’s systematic review concluded that smokers as a

group may not experience reduced fertility, but men with questionable sperm quality who
wish to have children may benefit from quitting smoking.

1

Commentary
One of the most substantial issues with many of the studies on this subject is the
problem of achieving the correct time-order between smoking exposure and its effect on
sperm quality. Determining that a particular exposure precedes the onset of an outcome
is one of the most important requirements of procuring a truly causal relationship. Most
of the studies reviewed were case-control or cross-sectional in design and it is impossible
to establish an inception cohort in both of these types of studies. In the case of crosssectional studies, both the exposure and the outcome are evaluated at the same time.
Comparably, in case-control studies, the outcome is evident, but the exposure status of
the case has to be obtained retrospectively. Essentially, neither study type is sufficient in
obtaining a correct time order between the exposure and the outcome. This is an obvious
weakness in the Chia et al. and Zhang et al. case-control studies on cigarette smoking and
10

sperm quality in which they selected cases and controls and retrospectively acquired the
participant’s exposure status.

28,29

Securing an inception cohort was also a problem in

Trummer, Sepaniak, and Kunzle’s cross-sectional studies on cigarette smoking and sperm
quality.

25, 26, 27

The studies all collected smoking exposures by administering

questionnaires close to collection of the semen samples. A prospective cohort is the
optimal study design in the attainment of an inception cohort. This type of study allows
the exposure to appear first, and the possible outcome to surface following the exposure.
This cause and effect relationship is the hallmark of a causal association.
Noticeable deficiencies in a number of studies regard the way in which smoking
exposure status was obtained. In a few of the studies smoking exposure was obtained by
simply “asking” the subject or by administering a questionnaire.

25,26,28,29

This method

of acquiring exposure status causes a definite complication: the subject could fabricate
his smoking status and provide a false answer, or he could miscalculate the number of
cigarettes smoked. Overall, this method could lead to bias and will not supply the
researchers with the most accurate exposure status. A more desirable way of obtaining
smoking exposure status was used by Sepaniak et al., in which subjects self-reported their
smoking status and also completed a CO test to confirm their smoking status.
31

32

27

Cotinine
33

level testing was also used in studies by Chen et al. , Wong et al. , Zenzes et al. , and
34

Sofikitis et al. ; all studies which found a significant association between cotinine
concentrations and semen parameters. Although CO testing is a simple improvement on
the sole use of questionnaires, it does not always correctly detect different levels of

11

smoking exposure. CO also has a very short half-life of 1-4 hours, so it would not be a
good measure of actual smoking exposure unless the subject smoked close to the testing
interval.

35

A better measurement of smoking exposure would be testing nicotine levels

in hair, as each centimeter of hair represents more than one month of exposure.

36

A number of studies that were reviewed did not fully describe the method in
which sperm parameter measurements were obtained or the way in which their samples
were analyzed. According to WHO guidelines, samples should be assessed within 1-2
hours of ejaculation.

37

For example, Trummer and Sofikitis do not mention the time

frame in which samples were analyzed.

30, 25

It is possible that analyzing samples after

the correct time interval could lead to loss of sample quality and this type of
measurement error may lead to a decrease in internal validity. The best course of action
in this case would be to analyze all samples within 1-2 hours after collection. This would
guarantee that the differences observed in the outcome measurements were not affected
by the time between collection and analysis. Additionally, many studies have failed to
mention who performed the sample analyses. It is important to know if one lab
technician analyzed the samples or if many technicians performed the task. Using
multiple technicians, assuming that there may be some extent of measurement error (as
they are different people), could in turn possibly lower the internal validity of the study.
Some studies also had a large window of time during which the semen samples
could be obtained. Such is the case in Kunzle et al.’s cross-sectional study in which
semen samples were obtained by ejaculation after 2-7 days of abstinence.

26

According to

WHO guidelines, semen samples should be collected after a minimum of 48 hours and no
12

37

longer than 7 days of sexual abstinence , but the possible differences in abstinence
length between subjects may decrease the internal validity of the study by introducing
measurement error. Although it may be difficult, an improvement may be to decrease the
time interval during which samples may be obtained from 2-7 days to 3-5 days of sexual
abstinence. This may improve the internal validity of the study by minimizing any
effects based on different lengths of abstinence.
Generally, strength of an association between cigarette smoking exposure and
sperm quality has not been demonstrated very well. In the studies that exhibited no
association, there was an obvious lack of evidence to make a significant causal inference
between the exposure and the outcome. However, a significant effect size was
demonstrated by Kunzle et al., which showed a 15.4%, 17.5% and 16.6% decrease in
sperm density, count, and motility, between smokers and non-smokers.

26

A large or

significant effect size is very helpful in establishing a causal association.
Another evident similarity between many studies has been the failure to
demonstrate a dose-response relationship between smoking and quality of semen
parameters. For instance, Chia, Trummer, Kunzle, and Sepaniak did not show a doseresponse relationship between the exposure and the outcome.

28, 25, 26, 27

However,

Zhang et al. did demonstrate a negative correlation between the amount and duration of
cigarette smoking and sperm density, viability, and forward progression.

29

Exhibiting a

dose-response relationship between an exposure and an outcome is very important in
achieving a causal association and makes the case for the causal relationship much
stronger.
13

Because of the many factors that are involved with smoking, the majority of
studies have not demonstrated great specificity. Since there are many other factors that
may cause harm to sperm quality, it is rather difficult to profess that cigarette smoking
exposure will undoubtedly have adverse effects on semen parameters. Cigarette smoking
has many detrimental outcomes and is therefore not specific to the outcome of sperm
quality. Specificity has also been very difficult to establish because of the lack of large
effect sizes and the problems with study design. If an exemplary study design is able to
show a large effect size, one could declare a high specificity in the magnitude of the
association.
A number of studies on this topic lacked external validity as much of the research
regarding semen parameters is accomplished at fertility clinics. The men who are usually
present at these clinics have reason to believe that their fertility may be compromised and
are therefore not a good generalization of the rest of the population. Although the motive
for gathering study cohorts from infertility clinics is understandable, it may compromise
the external validity of the study. In addition, most studies did not define ‘infertility’,
which may affect and contribute to the differences in results.
Finally, it should be pointed out that Vine’s systematic literature review did not
cover many of the studies reviewed for this thesis and thus may not be the best
1

representation of a recent systematic review. However, the results very do exemplify
the many disparate conclusions reached concerning the association of cigarette smoking
and sperm quality, as there is very low consistency on this issue. Some studies have
28, 29

found no association between the mentioned exposure and outcome
have found a significant association.

30, 25, 26, 29

14

, while others

As more studies on the subject matter

are completed, an agreeable consistency between study conclusions would aid in forming
a causal association.

15

METHODS
Objectives
This study reports on the effect of cigarette smoking on semen parameters from
data collected as part of the Fish and Infertility Study (FINS) project, a cross-sectional
NIH study undertaken to evaluate the impact of environmental factors on measures of
male infertility. Although FINS was not aimed toward the collection of cigarette smoke
exposure data, that information was collected as part of the environmental factors and
subsequently analyzed for this report. The five outcome measures evaluated were: total
sperm count, sperm concentration, volume, motility, and morphology. The primary
objective of this study was to compare the semen parameters of men who reported
exposure to cigarette smoking and those who did not by evaluating three smoking
variables: “Current Smoker”, “Ever Smoked”, and “Hours of Monthly Passive Smoking
Exposure”. Secondary objectives were to evaluate the effect of other lifestyle factors on
semen parameters of study subjects.
Study Design
Data collected for the NIH sponsored FINS study was analyzed for the purposes
of this report. The FINS project was undertaken to investigate the relationship between
measures of male reproductive health and exposure to environmental factors such as
organochlorines and other heavy metals. Subject data for the FINS study was collected
through self-reported questionnaires and included information regarding demographics
and other life style factors. Outcome variables were obtained through semen analyses.
This report uses the initial semen analysis parameters and relevant demographic and
exposure data, as reported in the FINS datasets.

16

Study Population
Prior to conducting the study, institutional review board (IRB) approval was
obtained from Michigan State University, Wayne State University, and the Michigan
Department of Community Health. Informed consent protocols gathered via the
mentioned institutions were followed in the recruitment and data collection from study
participants. Between 2002 and 2006, two infertility clinics in Michigan (University
Women’s Care (UWC), an affiliate of the Detroit Medical Center, Wayne State
University, Detroit, MI; and Grand Rapids Fertility and IVF, PC, in Grand Rapids, MI
were used as the location for subject recruitment. Men between the ages of 18 and 60
years old were recruited from couples presenting for infertility testing at the two
mentioned clinics. Invitations to participate in a project studying the impact of sportcaught fish consumption and other environmental factors on fertility were presented to all
men of couples presenting for infertility at the clinics. Men fitting the following medical
conditions were excluded for the study: diabetes, thyroid or adrenal disorder, genetic
disorders related to fertility, testicular cancer, unilateral orchiectomy, or use of hormonal
therapy. 603 men participated in the study.
Data Collection
Men who wished to participate in the study were properly enrolled using a
consent form and asked to provide semen, blood, and urine specimens. Each participant
was required to complete a detailed 50-page questionnaire administered by a trained
interviewer and study recruiter. The questionnaire inquired about basic demographics,
medical and occupational histories, and other factors of interest. Extensive information
on lifestyle and other environmental factors was also obtained and included the

17

following: smoking history, alcohol and caffeine consumption, underwear type, stress at
work and home, exercise, height, and weight.
Analysis of Semen Specimens
Each participant was asked to donate two semen specimens for analysis.
However, not all participants complied with this request. Semen analysis was performed
at the andrology laboratories at each study site and results were reported according to the
WHO guidelines.

38

Total sperm count, concentration, initial motility, Kruger’s strict

39

morphology , and semen volume were used in the statistical analyses. Considering the
non-uniformity of the number of semen specimen donations per participant, our analysis
only used the first semen analysis performed at each study site’s laboratory for each
subject. This sample commonly coincided with the semen analysis carried out closest to
the time of the questionnaire intake.
Data Analysis
Information from the questionnaire was summarized in Excel files and statistical
38

analysis was performed using SAS software (version 9.2). Using the WHO criteria , the
following semen parameters were classified as abnormal or low (per ejaculate): total
sperm count ≤ 40 million, sperm concentration ≤ 20 million/mL, motility ≤50% forward
progression, morphology ≤ 4% normal forms, and semen volume ≤ 2 mL (Table 1). Body
mass index was calculated from each participant’s reported height and weight using the
2

formula: BMI = (weight in Lbs * 703) / (height in inches) (Table 2). Three separate
smoking exposure variables were used to assess exposure in a unique and comprehensive
manner. The question “Do you smoke now?” was used to determine the current smoking
18

status of the subjects. In order to increase subject population in the smoking data, the
categorized variable “Ever Smoked” (1= Yes, 0= No) was created, merging any evidence
of past smoking into one inclusive variable. “Hours of monthly exposure passive
cigarette smoking” was categorized into three categories in order to examine a doseresponse effect. To contribute to a more meaningful analysis, all additional continuous
variables were turned into appropriate categorical variables, with exception of subject
weight, height, age (Table 3).
Each semen parameter was regarded as an outcome variable and dichotomized as
1= abnormal 0=normal using the WHO cut-offs. General linear modeling was used in
identifying difference in means of semen parameters between smoking variables and all
other demographics variables. Using logistic regression models and univariable analyses,
all variables, including all three smoke exposure variables, were individually assessed for
significance with each semen parameter. All variables with a p-values < 0.20 in the
univariable analyses were deemed significant and included in multivariable logistic
regression models for each of five separate semen outcomes and three smoke exposure
variables. Finally, full multivariable logistic regression models were applied to examine
the relationship between each smoke exposure variable and sperm parameter. All
multivariable logistic regression models were controlled for study site, education level,
annual income, age, race, and BMI, and factors with p< 0.05 were regarded as significant
(Table 4).

19

RESULTS
603 men participated in this study. The mean age of the participants was 34 years
old. The mean BMI was slightly below what is defined as obese (Table 5 and Figures 12). 74% of the subjects were Caucasian, 20% were African American, and 7% were
categorized as ‘other’ (Table 6). Almost half of the participants reported having a college
or post-college degree and nearly 30% reported having an annual income of > $90,000.
(Table 6) About half of the participants had previously fathered children; some through
assisted reproductive technologies and others on their own.
244 participants replied to a question framed as “Do you smoke now?” 49%
identified as current smokers and 51% identified as current non-smokers. The created
smoke exposure category “Ever Smoked” indicated 42% ‘Ever Smokers’ and 58%
‘Never Smokers’. With regard to second hand smoke exposure in the past month, 19% of
the respondents reported having no exposure, while 43% reported having < 10 hours and
38% reported having ≥10 hours (Table 7).
The minimum, maximum, median, mean, and standard deviation values for semen
parameters within the study population are listed in Table 8 and depicted in Figures 3-7.
Based on the WHO semen parameter guidelines we found the following frequency of
abnormal semen parameters within our population: 22% low total sperm count, 25% low
concentration, 34% low volume, 41% low motility, and 58% low normal morphology
(Table 9). In the general linear modeling, significant associations were found between
high second hand smoke exposure and low total sperm count (p=0.001) and low volume
(p=0.002) (Table 10). African Americans had significantly lower values of total sperm
count, concentration, and volume, compared to Whites and Others (Table 11). In general,

20

those who were obese had lower sperm concentration and percent normal motility than
those who were not. Those with less education had lower values of total sperm count
(p=0.025) and sperm concentration (p=0.017). However, those indicating their education
level as ‘high school graduate’ had a significantly higher percent normal morphology
than all others (p<0.001). Participants with higher income had generally higher values of
total sperm count, concentration, and volume, although not significantly. There was a
significant positive association between higher daily coffee consumption and increased
sperm concentration (p=0.033) (Table 12). Inversely, there was a significant negative
association between increased coffee consumption and percent normal motility
(p=0.024). Similar inconsistent associations were also found for daily hot tea, iced tea,
and pop. However, when caffeine intake was regarded as a whole, there were no
associations between weekly caffeine consumption and any of the semen parameters.
Our findings also suggested a significant association between increased monthly beer
consumption and higher semen count and concentration (p=0.014 and p=0.013,
respectively). Conversely, increased liquor consumption was associated with decreased
semen volume (p=0.035). When alcohol consumption was analyzed as a whole there
were no significant associations with any of the semen parameters (Table 13). Likewise,
there were no significant associations between the semen parameters, exercise, and
underwear type (Table 14). However, ‘no work stress’ and ‘moderate to severe work
stress’ were significantly associated with lower total sperm count (p=0.022), with those
experiencing ‘moderate to severe’ work stress having higher total sperm count levels.
The same pattern is seen in home stress, although not significantly (Table 14).

21

Using full logistic multivariable regression models, and including all variables
significant at P< 0.2 in the previous logistic univariable models, we found that current
and ever smoking status did not have any significant effect on low total sperm count
(Tables 15 and 16). Compared to those who reported zero hours of monthly passive
smoking, those who reported ≥10 hours of passive cigarette smoke exposure had 2.7
times the odds of having lower total sperm count (p=0.053) (Table 17). However, when
the model for low total sperm count was analyzed using current smoking and passive
smoking exposure as covariates, neither smoking exposure was significant (Table 18). In
general, work stress seemed to have a protective effect on total sperm count. Compared
to those who reported no stress at work, those who did experience work stress were much
less likely to have low total sperm count (Tables 15 and 18). Conversely, those who
reported exercising had more than twice the odds of having low total sperm count (Tables
15, 16, 17, and 18). We found no association between any of the smoke exposure
variables and low sperm concentration. Our results were mixed with regard to the effects
of work stress and exercise on low sperm concentration (Tables 19, 20, and 21).
Compared to those who reported zero hours of monthly passive smoking, those who
reported ≥10 hours of passive cigarette smoke exposure had 1.5 times the odds of having
lower semen volume (p=0.04) (Table 24). However, when the model for low semen
volume was analyzed using current smoking and passive smoking exposure as covariates,
neither smoking exposure was significant (Table 25). Overall, cigarette smoking
exposure did not seem to have any effect on semen volume or motility (Tables 22-28).
We did find that, compared to those who wore briefs, participants who reported wearing
boxers were significantly lower odds of having low sperm motility (Tables 26, 27, and

22

28). Finally, “Ever Smokers” were significantly more likely to have low normal sperm
morphology than “Never Smokers” (Table 30).

23

DISCUSSION
For this thesis, self-reported cigarette smoke exposure among men presenting at
two Michigan infertility centers was analyzed in order to assess its effects on five semen
parameters: total sperm count, concentration, volume, motility, and morphology. We
found a small, but significant, association between low normal sperm morphology in
participants who were “Ever Smokers” versus “Never Smokers”. No significant
associations were found between cigarette smoke exposure and sperm count,
concentration, semen volume, or motility. When total sperm count and sperm
concentration were assessed, exercise was shown to have a significant negative effect on
both semen parameters, while work stress seemed to function as a protective factor. We
also detected a protective association between the wearing of boxers and sperm motility.
Our study found a significant association between “Ever Smoked” status and low
normal sperm morphology. Study participants who had smoked were significantly more
likely to have low normal sperm morphology than those who reported never smoking.
Chia et al. demonstrated results along similar lines in their case-control study of 640
consecutive male partners of couples trying to conceive.

26

Cases referred to the male

partner of couples unable to conceive. Controls were defined as those of proven fertility
whose wives were pregnant at the time of providing the semen. The study found that
participants who smoked were 2.82 times more likely to be “infertile” than those who did
not. They concluded that participants who smoked cigarettes were significantly more
likely to have lower normal sperm morphology and in turn more likely to experience
infertility.

26

24

Unlike previous studies on the effects of cigarette smoke exposure and multiple
semen parameters, our analysis only identified significance between “Ever Smoking” and
decreased normal sperm morphology. Our analysis also examined the effects of passive
cigarette smoking, but did not find a significant association between second hand smoke
exposure and semen parameters. A recent second-hand smoke study in mice found that
sidestream tobacco smoke induces mutations in mouse sperm. Although the study did
not report specifically on semen count, it did suggest that paternal exposure to passive
cigarette smoke may have reproductive consequences.

40

To our knowledge, the effects

of passive cigarette smoke exposure on semen parameters have not been studied in
humans.
We also performed multivariable logistic modeling on two other smoking
variables to assess current smoking exposure and ever smoking exposure. The former
variable was available in our dataset. However, only 244 of our 603 study participants
had chosen to answer the question. In order to increase the chances of detecting an
association between cigarette smoke exposure and semen parameters, we decided to
increase the sample size of the smoke exposure variable by creating an additional
variable: “Ever Smoked”. The “Ever Smoked” variable was created by merging all
evidence of prior smoking into one inclusive variable. Subsequently, the new variable
increased our sample size and we discovered that 42% of our participants had smoked
sometime in their life and 58% had not.
Our univariable logistic modeling demonstrated a significant association between
“Current Smoking” and low semen volume and motility, and “Ever Smoked” and low
sperm morphology. However, the full multivariable models only showed a significant
25

association between ever smoking and decreased normal sperm morphology. The lack of
significance in the effect of current smoking status and semen quality is similar to that
found by a cross-sectional study by Trummer et al.

25

27

and a study by Sepaniak et al. .

Trummer’s assessment of 517 nonsmoker, 109 ex-smokers, and 478 smokers found that
there were no significant differences between the conventional semen parameters and the
smoke exposure status of their participants. Our results were similar to Trummer’s in a
number of ways. Like Trummer, we were able to analyze current smokers and nonsmokers. Moreover, our “Ever Smoked” variable was able to capture ex-smokers. Both
our study and Trummer’s found no significance between current smoking status and
semen parameters. However, contrary to Trummer’s results, our study analysis did
demonstrate a significant association between “Ever Smoking” exposure and decreasing
normal sperm morphology. Our findings were also similar to those from the study by
Sepaniak et al. Although our study population was much larger than that of Sepaniak’s
57 non-smokers and 51 smokers, our study design was the same and current smoking
status was self-reported. To Sepaniak’s benefit, however, a CO test was also used to
confirm participant’s cigarette smoke exposure. Overall, our results were similar in that
neither analysis identified a significant association between current smoking exposure
and semen parameters.
Our lack of significance between current smoking status and semen parameters
contrasted with two case-control studies by Zhang et al. and Chia et al. The former study
revealed that current smoking among 110 non-smokers and 191 smokers did affect semen
quality and concluded that, compared to non-smokers, semen volume, density, viability,
and forward progression was significantly lower in medium, heavy, and long-term
26

smokers.

29

Chia et al. also found that current cigarette smoking significantly affected all

semen parameters and that the risk of infertility was associated with smoking.

28

Both

studies differed from ours in study design and results, but were similar in obtaining
current smoking status through questionnaire.
The multivariable logistic modeling also revealed a number of other significant
associations between exposure variables and semen parameters. We found that,
compared to those who reported experiencing “no work stress”, those who did were
significantly less likely to have low total sperm count and concentration. When analyzed
by annual income, our results revealed that as annual income increased, so did work
stress. However, given our initial univariable analysis of mean semen parameters, we
discovered that those with higher annual income were generally more likely to have
increased total sperm count, concentration, and volume. We propose that work stress is
simply a mediator variable that explains the relationship between income and semen
parameters. i.e. those men with higher paying, more secure jobs may have less work
stress than those with lower paying jobs. In addition, those who reported having “no
work stress” may have been without a job and answered the question as having no stress
at work.
Our analysis indicated that, compared to those who did not exercise, those who
reported exercising were much more likely to suffer from low total sperm count and
concentration. Literature on exercise and its effect on semen parameters are mixed and
have mostly relied on very small populations. One study on 10 long distance competitive
cyclist and 10 controls found that cyclists had significantly lower percent normal
morphology, but did not see a significant difference in other semen parameters.
27

41

Another study found that, compared to controls, some endurance-trained subjects had
reduced gonadotropin releasing hormones and lower total sperm count, decreased
motility, and decreased percent normal morphology.

42

Most recently, a prospective

cohort study of 2,261 men attending one of three IVF clinics in the Boston area during
1993-2003, found that none of the semen parameters were significantly associated with
regular exercise. However, compared with no regular exercise, bicycling ≥5 hours per
week was associated with low sperm concentration and low motility (OR 1.92 and 2.05,
respectively).

43

Because most studies on this subject, with exception of our study and

the mentioned recent prospective study, had a very small number of participants, they
may not have enough power to detect significant association, or to come to meaningful
conclusions.
Finally, our multivariable modeling did reveal a significant association between
underwear type and sperm motility. Compared to those who wore briefs, those who wore
boxers were significantly less likely to have low sperm motility. This seems to support
the theory of tight fitting undergarments, high scrotal temperature, and its negative
impact on various semen parameters. Similar to studies on exercise and semen
parameters, literature on the impact of underwear type on sperm is mixed. One pilot
study evaluated the effect of fit of underwear on sperm production in two healthy males.
The participants alternated wearing tight fitting briefs and loose fitting boxers in an
ABAB withdrawal design study in which the conditions lasted three months and were
alternated twice to results in a one year study. The results showed that semen parameters
gradually decreased in tight conditions and increased in loose conditions.

28

44

Another

case-crossover study in 97 consecutive men presenting for evaluation of primary clinical
subfertility found no significant association between underwear type, scrotal
temperatures, and semen parameters.

40

Still another case-control study found that men

who wore tight underpants or trousers were significantly more likely to present with
dyspermia (OR 1.9 and 1.6, respectively).

45

This study has a number of strengths that deserve mention. The dataset used for
this analysis contained a large number of individuals with complete data on a number of
lifestyles factors including demographics, caffeine and alcohol intake, exercise, and most
importantly cigarette smoking habits. The large sample size increased the statistical
power of our analyses and the comprehensive lifestyle factors allowed us to create
complex statistical models, utilizing the variety of exposure variables and the five semen
outcome variables. The multivariable logistic regression models enabled us to evaluate
multiple risk factors, while allowing for control of confounding variables. In addition,
our dataset included very complete information on hours of monthly passive cigarette
smoke exposure, a factor that has rarely been studied. Moreover, because of the variety
of questions on cigarette smoke exposure, we were able create a third variable to assess
participant’s history of ever smoking.
This study does have some limitations. The first is that the study population
consisted of men presenting to two infertility clinics. This study population is not
representative of normal healthy men and the use of this population could limit the
generalizability of the findings and lead to external validity issues. However, about half
of the subjects had fathered children previously. The design of this study was crosssectional, which may lead to some inherent time-order issues between the exposure
29

variables and the outcome measures. Additionally, all study data, sans the semen
parameters, were obtained through self-reporting on lengthy questionnaires which may
result in recall bias. The dataset used for this study was not designed for an analysis of
cigarette smoking exposure and semen parameters, and less than half of the total study
population revealed their current smoking status. Even then, we cannot be sure of the
presence of bias in their responses associated with stigma of cigarette smoking. Indeed,
we created a third smoke exposure variable to assess “Ever Smoking” in order to obtain a
larger number of responses on cigarette smoke exposure. Lastly, although we attempted
to be as consistent as possible, the cut-off points and categorizations of some lifestyle
factors were set arbitrarily based on our data points. It is reasonable to assume that
different categorizations or cut-offs may have the possibility of leading to other results.
In summary, this study analyzed the effects of cigarette smoke exposure on five
semen parameters. We found that those who reported ever smoking had higher odds of
low normal sperm morphology (OR=1.61, p=0.032, CI= 1.043, 2.496). Our analysis also
revealed trends for a protective effect of work stress and low total sperm count and
concentration, and also of boxer wear and low sperm motility. Lastly, we suggest the
implementation of future prospective cohort studies with large study populations geared
toward studying cigarette smoke exposure, namely passive exposure, and its effects on
semen parameters.

30

APPENDIX

31

Table 1. Classification of low or abnormal semen parameters.
Semen Parameter
Total Sperm Count
Sperm Concentration
Semen Volume
Percent Motile Sperm
Percent Morphologically Normal Sperm
*

According to WHO criteria.

*

Considered Low/Abnormal
≤ 40 Million
≤ 20 Million/mL
≤ 2 mL
≤ 50% Forward Progression
≤ 4% Normal Forms

31

32

Table 2. Body Mass Index (BMI) categories and calculation.
BMI

Weight Category

18.5-24.99
25.0-29.99
30.0-39.99

Not
Overweight/Obese
Overweight
Obese

≥ 40.00

Morbidly Obese
BMI= weight (Lb)*703
2

height (in)

33

Table 3. Variables used in analyses, by unit and type.
Variable
Study Site
Age
Weight
Height
BMI
Race
Education
Annual Income
Smoke Now
Ever Smoked
Second Hand Smoke
Exposure
Total Sperm Count
Sperm Concentration
Semen Volume
Percent Motile Sperm
Percent Morphologically
Normal Sperm
Low Total Sperm Count
Low Sperm Concentration
Low Semen Volume
Low Motile Sperm
Low Normal Morphology
Smoke Now
Ever Smoked
Coffee Consumption
Hot Tea Consumption
Iced Tea Consumption
Pop Consumption
Weekly Caffeine Intake
Beer Consumption
Wine Consumption
Liquor Consumption
Mixed Drink
Consumption
Weekly Alcohol Intake
Exercise
Home Stress
Work Stress
Underwear Type

Units
Detroit/Grand Rapids
Numerical
Lb
Inches
Normal-Morbidly Obese
White/Black
Education Level
U.S. Dollars
Yes/No
Yes/No
Hours/Month

Type of Variable
Categorical
Continuous
Continuous
Continuous
Continuous
Categorical

No. of Sperm per Ejaculate
No. of Sperm/mL of
Ejaculate
mL
Percent (%)
Percent (%)

Continuous
Continuous

1/0
1/0
1/0
1/0
1/0
Yes/No
Yes/No
Cups/Day
Cups/Day
Cups/Day
12 OZ Cans/Day
None-High
Beers/Month
Glasses of Wine/Month
Shots of Liquor/Month
No. Mixed Drinks/Month

Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical
Categorical

None-High
Yes/No
None-Severe
None-Severe
Boxers/Briefs/Other

Categorical
Categorical
Categorical
Categorical
Categorical

34

Categorical
Categorical
Categorical
Categorical

Continuous
Continuous
Continuous

Table 4. Outcomes and exposure used in multivariable logistic regression models.
Model Name

Outcome
Variable

1a: Low Total Sperm Count and Smoke Now
1b: Low Total Sperm Count and Ever Smoked
1c: Low Total Sperm Count and Second Hand
Smoke Exposure

Low total
sperm count

1d: Low Total Sperm Count, Smoke Now,
and Second Hand Smoke Exposure
2a: Low Sperm Concentration and Smoke
Now
2b: Low Sperm Concentration and Ever
Smoked

Low sperm
concentration

2c: Low Sperm Concentration and Second
Hand Smoke Exposure

3a: Low Semen Volume and Smoke Now
3b: Low Semen Volume and Ever Smoked
3c: Low Semen Volume and Second Hand
Smoke Exposure
3d: Low Semen Volume, Second Hand
Smoke Exposure, and Smoke Now
4a: Low Sperm Motility and Smoke Now

Low semen
volume

Low sperm
motility

4b: Low Sperm Motility and Ever Smoked
4c: Low Sperm Motility and Second Hand
Smoke Exposure
5a: Low Normal Sperm Morphology and
Smoke Now
5b: Low Normal Sperm Morphology and Ever
Smoked
5c: Low Normal Sperm Morphology and
Second Hand Smoke Exposure
*

Low normal
sperm
morphology

*

Exposure Variables
Smoke now, home stress,
work stress, exercise
Ever smoked, home stress,
work stress, exercise
Second hand smoke
exposure, home stress, work
stress, exercise
Smoke now, second hand
smoke exposure, home
stress, work stress, exercise
Smoke now, home stress,
work stress, exercise,
weekly caffeine intake
Ever smoked, home stress,
work stress, exercise,
weekly caffeine intake
Second hand smoke
exposure, home stress, work
stress, exercise, weekly
caffeine intake
Smoke now
Ever smoked
Second hand smoke
exposure
Second hand smoke
exposure, smoke now
Smoke now, underwear
type
Ever smoked, underwear
type
Second hand smoke
exposure, underwear type
Smoke now, work stress
Ever smoked, work stress
Second hand smoke
exposure, work stress

All models were controlled for study site, education level, annual income, age, race, and
BMI. Factors with P <0.05 were regarded as significant.

35

Table 5. Demographic characteristics of participants.
Demographic
Characteristics
Study Site
Grand Rapids
Detroit
Age
Weight (Lb)
Height (In)
BMI

N

Min

208
393
594
602
603
602

18
123
62
16.68

Max

Median

Mean

Standard
Deviation

60
465
83
55.13

34
200
71
27.82

34.13
208.85
70.66
29.33

5.85
46.06
3.07
5.88

36

Table 6. Socio-economic characteristics of participants.
Socio-Economic
Characteristics
Race
White
Black
Other
Education
Elementary
Some High School
High School Graduate
Some College
College or PostCollege Graduate
Annual Income
≤$29,999
$30,000- ≤$44,999
$45,000- ≤$59,999
$60,000- ≤$74,999
$75,000- ≤$89,999
≥$90,000

N

Frequency

%

437
120
43

73.83
20.00
7.17

7
31
97
200
268

1.16
5.14
16.09
33.17
44.44

70
69
99
86
83
167

12.20
12.02
17.25
14.98
14.46
29.09

600

603

574

37

Table 7. Smoking characteristics of participants.
Smoking Exposure
Smoke Now
a
No

N
244

Yes
Ever smoked
c
No

50.82

124

49.18

350

58.04

253

41.96

112
259
227

18.73
43.31
37.96

603

d

Yes
Second Hand
Smoke (Monthly)
None
<10 Hours
≥10 Hours

%

120

b

Frequency

598

a Replied “No” to “Do you smoke now?”

b Replied “Yes” to “Do you smoke now?”

c Did not reply “Yes” to any of the following questions: “Have you smoked >100

cigarettes (5 packs) in your lifetime?”, “Have you smoked cigarettes in the last twelve
months?”, or “Do you smoke now?”

d Replied “Yes” to any of the following questions: “Have you smoked >100 cigarettes (5

packs) in your lifetime?”, “Have you smoked cigarettes in the last twelve months?”, and
“Do you smoke now?”

38

Table 8. Semen parameters of participants.
Semen
Parameters
Total Sperm
6
Count (10 )
Sperm
Concentration
6
(10 /mL)
Semen Volume
(mL)
Motile Sperm
(%)
Morphologically
Normal Sperm
(%)

Standard
Deviation

N

Min

Max

Median

Mean

575

0

948.50

117.25

159.92 149.48

575

0

364.00

43.80

59.27

54.50

577

0.02

12.00

2.90

2.93

1.52

571

0

93.00

54.00

51.35

16.16

535

0

22.00

4.00

4.60

4.02

39

Table 9. Semen parameters of participants classified by “Normal” or “Abnormal”.
Semen Parameters
6

Total Count (10 )
Normal
a
Abnormal
Concentration
6
(10 /mL)
Normal
b
Abnormal
Volume (mL)
Normal
c
Abnormal
Motility (%)
Normal
d
Abnormal
Normal
Morphology (%)
Normal
e
Abnormal
*

N
575

Frequency

%

451
124

78.43
21.57

433
142

75.30
24.70

381
196

66.03
33.97

340
231

59.54
40.46

224
311

41.87
58.13

575

577

571

535

According to WHO criteria.

31

a

Total sperm count ≤ 40 million was considered abnormal

b

Sperm concentration ≤ 20 million/mL was considered abnormal

c

Semen volume ≤ 2 mL was considered abnormal

d

Sperm motility ≤ 50% forward progression was considered abnormal

e

Sperm morphology ≤ 4% normal forms was considered abnormal

40

*

Table 10. Univariable analysis depicting mean semen parameters and cigarette
smoke exposure.
Smoking
Exposure

Concentration
6
(10 /mL)

Volume
(mL)

Motility
(%)

Normal
Morphology
(%)

120 143.1

52.51

2.87

50.41

4.48

Ye s
P-Value
Ever
Smoked
c
No

124 133.2

57.95

2.66

51.79

4.79

0.555

0.422

0.248

0.505

0.570

350 159.5

58.18

3.00

51.18

4.50

d

235 160.4

60.73

2.83

51.58

4.72

0.943

0.577

0.209

0.767

0.533

59.79
63.67
53.84
0.150

2.86
3.19
2.70
0.002

50.04
51.53
51.79
0.639

5.04
4.52
4.48
0.479

Smoke Now
a
No
b

Ye s
P-Value
Second
Hand
Smoke
(Monthly)
None
<10 Hours
≥10 Hours
P-Value
a
b

N

Total
Sperm
Count
6
(10 )

112 158.51
259 183.99
227 133.69
0.001

Replied “No” to “Do you smoke now?”

Replied “Yes” to “Do you smoke now?”

Replied “Yes” to any of the following questions: “Have you smoked >100 cigarettes (5
packs) in your lifetime?”, “Have you smoked cigarettes in the last twelve months?”, or
“Do you smoke now?”

c

Did not reply “Yes” to any of the following questions: “Have you smoked >1000
cigarettes (5 packs) in your lifetime?”, “Have you smoked cigarettes in the last twelve
months?”, and “Do you smoke now?”

d

41

Table 11. Univariable analysis depicting mean semen parameters and subject
demographics.
Demographic
N
Characteristics

Race
White
Black
Other
P-Value
Age
18-29
30-39
40-60
P-Value
BMI
Normal
Overweight
Obese
Morbidly
Obese
P-Value
Education
Some High
School
High School
Grad
Some College
College Grad
P-Value
Annual
Income
<$29,999
$30,000-44,999
$45,000-59,999
$60,000-74,999
$75,000-89,999
>$90,000
P-Value

Total
Sperm
Count
6
(10 )

Concentration
6
(10 /mL)

Volume
(mL)

Motility
(%)

Normal
Morphology
(%)

437 172.02
120 111.51
43 168.44
<0.001

62.08
48.34
59.55
0.061

3.08
2.39
2.94
<0.001

51.31
51.22
52.29
0.934

4.70
4.66
3.29
0.135

128 151.08
371 169.99
95 133.46
0.094

55.70
55.59
49.13
0.419

2.90
3.01
2.69
0.211

15.74
16.59
14.80
0.076

4.93
4.79
3.46
0.016

123
262
180
37

180.93
157.98
161.25
104.31

68.45
57.59
59.96
40.10

2.88
2.97
2.92
2.90

49.55
51.50
53.47
46.14

4.28
4.72
4.80
3.91

0.064

0.047

0.962

0.043

0.520

31

147.81

59.66

2.74

52.66

4.10

97

127.64

49.92

2.67

52.53

6.16

200 151.22
268 179.37
0.025

53.19
67.03
0.017

3.02
2.99
0.236

50.40
51.27
0.734

3.96
4.62
<0.001

70
69
99
86
83
167

49.59
50.65
63.31
67.53
60.56
62.25
0.269

2.56
3.00
2.84
2.90
3.29
3.03
0.088

53.58
52.53
51.15
50.21
50.68
51.31
0.835

4.27
4.52
5.41
5.17
4.15
4.42
0.234

125.37
143.41
156.49
184.62
187.42
165.96
0.097

42

Table 12. Univariable analysis depicting mean semen parameters and caffeine
consumption.
Caffeine
Consumption

Concentration
6
(10 /mL)

Volume
(mL)

Motility
(%)

Normal
Morphology
(%)

240 148.64
176 156.83
182 179.30
0.115

54.81
56.46
68.26
0.033

2.87
2.98
2.98
0.699

53.12
51.90
48.74
0.024

4.86
4.20
4.72
0.272

459 164.12
100 125.01
38 199.60
0.019

61.17
46.67
68.98
0.036

2.96
2.77
3.02
0.509

52.00
48.44
52.11
0.154

4.83
4.01
3.58
0.066

399 175.12
142 139.20
56 110.77
0.002

63.83
53.56
43.40
0.013

2.97
2.88
2.86
0.785

52.06
49.62
51.58
0.321

4.68
4.28
4.98
0.522

94 168.24
220 157.35
287 159.95
0.836

58.18
59.52
59.48
0.978

2.93
3.01
2.88
0.629

49.46
52.31
51.20
0.378

4.44
4.55
4.65
0.905

21

184.75

65.52

2.49

53.05

4.17

114 145.86

52.59

3.00

51.10

4.63

c

39

130.98

47.09

3.20

52.16

3.62

d

61

156.07

56.12

2.84

53.00

5.28

338 166.74

63.14

2.94

51.07

4.58

0.236

0.540

0.906

0.415

Daily Coffee
None
<2 cups
≥ 2 cups
P-Value
Daily Hot Tea
None
<1 cups
≥ 1 cups
P-Value
Daily Iced Tea
None
<1/2 cups
≥ 1/2 cups
P-Value
Daily Pop
None
<1 pop
≥ 1 pop
P-Value
Weekly
Caffeine
Intake
a
None
Low

b

Low-Mod
Moderate
e

High
P-Value
a
b
c
d
e

N

Total
Sperm
Count
6
(10 )

0.469

No caffeine consumption per week
1 to < 7 caffeinated drinks per week
7 to < 14 caffeinated drinks per week
14 to < 50 caffeinated drinks per week
≥ 50 caffeinated drinks per week

43

Table 13. Univariable analysis depicting mean semen parameters and alcohol
consumption.
Alcohol
Consumption

Monthly Beer
None
<30 beers
≥ 30 beers
P-Value
Monthly Wine
None
<10 glasses
≥ 10 glasses
P-Value
Monthly
Liquor
None
<3 shots
≥ 3 shots
P-Value
Monthly
Mixed Drinks
None
<4 drinks
≥ 4 drinks
P-Value
Weekly
Alcohol Intake
a
None
Low

b

Low-Mod
Moderate
e

High
P-Value
a
b
c
d
e

N

Total
Sperm
Count
6
(10 )

Concentration
6
(10 /mL)

Volume
(mL)

Motility
(%)

Normal
Morphology
(%)

199 134.39
331 170.30
73 181.71
0.014

51.09
61.30
71.91
0.013

2.83
3.02
2.82
0.346

49.22
52.31
52.65
0.091

4.40
4.50
5.50
0.127

397 154.00
168 175.62
34 155.39
0.302

57.48
63.27
61.81
0.510

2.88
3.06
2.91
0.457

51.87
50.67
50.50
0.689

4.69
4.67
3.61
0.359

420 164.73
86 151.79
94 145.19
0.469

59.14
57.59
61.12
0.913

3.03
2.89
2.57
0.035

51.41
51.90
50.61
0.866

4.72
4.59
4.18
0.538

372 155.66
118 177.63
109 154.34
0.359

59.79
58.71
58.21
0.959

2.87
3.18
2.88
0.153

51.58
49.18
52.85
0.217

4.62
4.41
4.75
0.829

132 136.77

51.35

2.92

50.02

4.25

116 164.32

61.58

2.93

50.79

4.42

c

131 170.31

61.84

3.03

52.79

4.71

d

89

181.56

64.67

2.85

50.49

4.73

43

140.28

63.59

2.56

52.59

4.82

0.179

0.380

0.538

0.661

0.861

No alcoholic drink consumption per week
1 to < 3 alcoholic drinks per week
3 to < 7 alcoholic drinks per week
7 to < 25 alcoholic drinks per week
≥ 25 alcoholic drinks per week
44

Table 14. Univariable analysis depicting mean semen parameters and other lifestyle
factors.
Lifestyle
Factors

Exercise
Yes
No
P-Value
Home
Stress
None
Slight- Mod
ModSevere
P-Value
Work
Stress
None
Slight- Mod
ModSevere
P-Value
Underwear
Type
Boxers
Briefs
Other
P-Value

N

Total
Sperm
Count
6
(10 )

Concentration
6
(10 /mL)

Volume
(mL)

Motility
(%)

Normal
Morphology
(%)

250
270

160.24
167.30
0.608

56.88
63.84
0.161

3.03
2.85
0.196

51.29
52.25
0.512

4.69
4.47
0.545

67
415
117

151.89
168.99
132.38

56.88
62.41
49.91

3.02
2.93
2.85

49.35
51.64
51.56

5.03
4.68
4.07

0.065

0.092

0.773

0.574

0.263

123.59
177.33
150.71

47.97
64.00
58.12

2.82
3.00
2.88

53.22
52.13
51.26

5.04
4.75
4.46

0.022

0.116

0.557

0.690

0.608

169.68
155.69
165.73
0.835

60.15
57.71
60.79
0.856

2.92
2.90
2.95
0.954

52.60
50.23
50.24
0.275

4.67
4.57
4.49
0.921

53
323
199

254
190
152

45

Table 15. Logistic regression model for Low Total Sperm Count, using “Smoke
*
Now” as the main variable (N=145).
Variable

Odds Ratio

95% Confidence
Interval

Smoke Now
No

Ref.

Yes

1.23

(0.416, 3.647)

Ref.
2.70
3.28

(0.355, 20.591)
(0.366, 29.422)

Ref.
0.08
0.07

(0.013, 0.440)
(0.010, 0.489)

Ref.
3.50

(1.197, 10.260)

P-Value

0.707

Home Stress
None
Slight- Mod
Mod- Severe
Work Stress
None
Slight- Mod
Mod- Severe
Exercise
No
Yes
*

0.566

0.014

0.022

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

46

Table 16. Logistic regression model for Low Total Sperm Count, using “Ever
*
Smoked” as the main variable (N=448).
Variable

Odds Ratio

Ever Smoked
No

95% Confidence
Interval

P-Value

Ref.
0.945

Yes
Home Stress
None
Slight- Mod
Mod- Severe
Work Stress
None
Slight- Mod
Mod- Severe
Exercise
No
Yes

1.02

(0.591, 1.758)

Ref.
1.68
2.48

(0.673, 4.186)
(0.876, 7.009)

0.215

Ref.
0.38
0.38

(0.150, 0.966)
(0.141, 1.040)

0.118

Ref.
2.63

(1.489, 4.632)

*

0.001

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

47

Table 17. Logistic regression model for Low Total Sperm Count, using “Monthly
*
Hours of Passive Smoking” as the main variable (N=445).
Variable

Odds Ratio

Hours of Passive
Smoking (Monthly)
None

Ref

< 10 hours
≥ 10 hours
Home Stress
None
Slight- Mod
Mod- Severe
Work Stress
None
Slight- Mod
Mod- Severe
Exercise
No
Yes

95% Confidence
Interval

P-Value

1.70

(0.722, 4.000)

0.053

2.69

(1.163, 6.222)

Ref.
1.64
2.37

(0.652, 4.118)
(0.828, 6.763)

0.259

Ref.
0.40
0.37

(0.152, 1.050)
(0.133, 1.038)

0.147

Ref.
2.55

(1.437, 4.540)

0.001

*

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

48

Table 18. Logistic regression model for Low Total Sperm Count, using “Monthly
*
Hours of Passive Smoking” and “Smoke Now” as main variables (N=143).
Variable

Odds Ratio

Hours of Passive
Smoking (Monthly)
None

Ref

< 10 hours
≥ 10 hours
Smoke Now
No
Yes
Home Stress
None
Slight- Mod
Mod- Severe
Work Stress
None
Slight- Mod
Mod- Severe
Exercise
No
Yes

95% Confidence
Interval

P-Value

7.58

(0.595, 96.502)

0.259

4.11

(0.370-45.726)
0.644

Ref.
1.30

(0.424, 3.997)

Ref.
2.37
2.44

(0.321, 17.583)
(0.273, 21.819)

0.691

Ref.
0.08
0.09

(0.012, 0.613)
(0.012, 0.470)

0.021

Ref.
3.40

0.028
(1.138, 10.146)

*

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

49

Table 19. Logistic regression model for Low Sperm Concentration, using “Smoke
*
Now” as the main variable (N=141).
Variable

Odds Ratio

Smoke Now
No

95% Confidence
Interval

P-Value

Ref.
0.735

Yes
Home Stress
None
Slight- Mod
Mod- Severe
Work Stress
None
Slight- Mod
Mod- Severe
Exercise
No
Yes
Weekly Caffeine
Intake
None
Low
Low-Mod
Moderate
High

1.21

(0.405, 3.604)

Ref.
1.08
2.20

(0.119, 9.788)
(0.199, 24.221)

0.485

Ref.
0.04
0.05

(0.006, 0.306)
(0.007, 0.425)

0.007

0.150

Ref.
2.22

(0.762, 5.935)

Ref.
0.19
0.03
<0.001
0.07

(0.004, 8.445)
(<0.001, 2.408)
(<0.001, >999.999)
(0.001, 3.074)

*

0.424

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

50

Table 20. Logistic regression model for Low Sperm Concentration, using “Ever
*
Smoked” as the main variable (N=425).
Variable

Odds Ratio

Ever Smoked
No

95% Confidence
Interval

P-Value

Ref.
0.960

Yes
Home Stress
None
Slight- Mod
Mod- Severe
Work Stress
None
Slight- Mod
Mod- Severe
Exercise
No
Yes
Weekly Caffeine
Intake
None
Low
Low-Mod
Moderate
High

1.01

(0.589, 1.745)

Ref.
1.37
2.45

(0.563, 3.316)
(0.891, 6.747)

0.122

Ref.
0.22
0.28

(0.085, 0.549)
(0.104, 0.742)

0.006

< 0.001

Ref.
2.87

(1.652, 4.992)

Ref.
1.13
1.20
0.59
0.94

(0.318, 4.042)
(0.253, 5.667)
(0.140, 2.485)
(0.275, 3.185)

*

0.771

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

51

Table 21. Logistic regression model for Low Sperm Concentration, using “Monthly
*
Hours of Passive Smoking” as the main variable (N= 422).
Variable

Odds Ratio

Hours of Passive
Smoking (Monthly)
None
< 10 hours
≥ 10 hours
Home Stress
None
Slight- Mod
Mod- Severe
Work Stress
None
Slight- Mod
Mod- Severe
Exercise
No
Yes
Weekly Caffeine
Intake
None
Low
Low-Mod
Moderate
High

95% Confidence
Interval

P-Value

Ref
1.35

(0.599, 3.025)

0.464

1.66

(0.734, 3.757)

Ref.
1.33
2.48

(0.549, 3.237)
(0.898, 6.827)

0.108

Ref.
0.23
0.29

(0.090, 0.604)
(0.106, 0.783)

0.011

Ref.
0.86

(1.575, 4.777)

Ref.
1.15
1.44
0.62
0.89

(0.323, 4.104)
(0.301, 6.902)
(0.147, 2.583)
(0.263, 2.983)

*

<0.001

0.694

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

52

Table 22. Logistic regression model for Low Semen Volume, using “Smoke Now” as
*
the main variable (N=216).
Variable

Odds Ratio

Smoke Now
No

Ref.

Yes

95% Confidence
Interval

1.66

P-Value

0.132
(0.858, 3.209)

*

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

53

Table 23. Logistic regression model for Low Semen Volume, using “Ever Smoked”
*
as the main variable (N=535).
Variable

Odds Ratio

Ever Smoked
No

Ref.

Yes

95% Confidence
Interval

0.99

P-Value

0.984
(0.671, 1.478)

*

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

54

Table 24. Logistic regression model for Low Semen Volume, using “Monthly Hours
*
of Passive Smoking” as the main variable (N=531).
Variable

Odds Ratio

95% Confidence
Interval

P-Value

Hours of Passive
Smoking (Monthly)
None
< 10 hours

Ref
0.81

(0.467, 1.422)

0.039

≥ 10 hours

1.48

(0.850, 2.583)

*

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

55

Table 25. Logistic regression model for Low Semen Volume, using “Monthly Hours
*
of Passive Smoking” and “Smoke Now” as the main variables (N=213).
Variable

Odds Ratio

Hours of Passive
Smoking (Monthly)
None
< 10 hours

95% Confidence
Interval

Ref
0.78

P-Value

(0.272, 2.243)

0.843

≥ 10 hours
0.97
(0.362, 2.571)
Smoke Now
No
Ref.
0.283
Yes
1.46
(0.733, 2.903)
*
Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

56

Table 26. Logistic regression model for Low Sperm Motility, using “Smoke Now” as
*
the main variable (N=150).
Variable

Odds Ratio

Smoke Now
No

95% Confidence
Interval

P-Value

Ref.
0.300

Yes
Underwear Type
Briefs
Boxers
Other

0.63

(0.268, 1.501)

Ref.
0.15
0.63

(0.050, 0.424)
(0.228, 1.727)

*

0.001

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

57

Table 27. Logistic regression model for Low Sperm Motility, using “Ever Smoked”
*
as the main variable (N=461).
Variable

Odds Ratio

Ever Smoked
No

95% Confidence
Interval

P-Value

Ref.
0.874

Yes
Underwear Type
Briefs
Boxers
Other

1.04

(0.673, 1.593)

Ref.
0.55
0.96

(0.335, 0.901)
(0.562, 1.645)

*

0.028

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

58

Table 28. Logistic regression model for Low Sperm Motility, using “Monthly Hours
*
of Passive Smoking” as the main variable (N=458).
Variable

Odds Ratio

Hours of Passive
Smoking (Monthly)
None
< 10 hours
≥ 10 hours
Underwear Type
Briefs
Boxers
Other

95% Confidence
Interval

P-Value

Ref
0.75

(0.408, 1.362)

0.517

0.95

(0.519, 1.748)

Ref.
0.57
0.97

(0.345, 0.926)
(0.565, 1.653)

*

0.039

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

59

Table 29. Logistic regression model for Low Normal Sperm Morphology, using
*
“Smoke Now” as the main variable (N=212).
Variable

Odds Ratio

Smoke Now
No

95% Confidence
Interval

P-Value

Ref.
0.505

Yes
Work Stress
None
Slight- Mod
Mod- Severe

1.27

(0.628, 2.577)

Ref.
0.71
0.59

(0.217, 2.327)
(0.167, 2.100)

*

0.708

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

60

Table 30. Logistic regression model for Low Normal Sperm Morphology, using
*
“Ever Smoked” as the main variable (N= 530).
Variable

Odds Ratio

Ever Smoked
No

95% Confidence
Interval

P-Value

Ref.
0.032

Yes
Work Stress
None
Slight- Mod
Mod- Severe

1.61

(1.043, 2.496)

Ref.
0.83
0.67

(0.287, 1.755)
(0.302, 1.502)

*

0.548

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

61

Table 31. Logistic regression model for Low Normal Sperm Morphology, using
*
“Monthly Hours of Passive Smoking” as the main variable (N=526).
Variable

Odds Ratio

Hours of Passive
Smoking (Monthly)
None
< 10 hours
≥ 10 hours
Work Stress
None
Slight- Mod
Mod- Severe

95% Confidence
Interval

P-Value

Ref.
0.65

(0.359, 1.182)

0.269

0.62

(0.338, 1.147)

Ref.
0.84
0.70

(0.391, 1.792)
(0.314, 1.576)

*

0.631

Model was controlled for study site, education level, annual income, age, race, and BMI.
Factors with P <0.05 were regarded as significant.

62

Figure 1. Age of study subjects.

*

22.5
20.0
17.5
15.0
Percent

12.5
10.0
7.5
5.0
2.5
0
18

21 24

27 30

33

36 39

42

45 48

51

54 57

60

Age
*

For interpretation of the references to color in this and all other figures, the reader is
referred to the electronic version of this thesis.

63

Figure 2. Body Mass Index (BMI) of study subjects.
30
25

Percent

20
15
10
5
0
18

21

24

27

30

33

36

BMI

64

39

42

45

48

51

54

6

Figure 3. Total sperm count of study subjects (10 ).
40
35

Percent

30
25
20
15
10
5
0
40

120

200 280 360

440 520

600 680 760

Total Sperm Count

65

840 920

6

Figure 4. Semen concentration of study subjects (10 /mL).
35
30
25

Percent

20
15
10
5
0

0

30

60

90

120 150

180 210 240 270 300 330 360

Sperm Concentration (Sperm/mL)

66

Figure 5. Ejaculate volume of study subjects (mL).
35
30

Percent

25
20
15
10
5
0
0

1

2

3

4

5

6

7

8

Volume of Ejaculate (mL)

67

9

10

11

12

Figure 6. Percent sperm motility of study subjects.
30
25

Percent

20
15
10
5
0
4

12

20

28

36

44

52

Semen Motility

68

60

68

76

84

92

Figure 7. Percent normal sperm morphology of study subjects.
25

Percent

20
15
10
5
0
0

2

4

6

8

10

12

14

Semen with Normal Forms

69

16

18

20

22

BIBLIOGRAPHY

70

BIBLIOGRAPHY

1

Vine MF. Smoking and male reproduction: a review. International Journal of
Andrology. 1996; 19: 323-337.

2

U.S. Department of Health and Human Services, Centers for Disease Control and
Prevention, National Center for Chronic Disease Prevention and Health Promotion,
Office on Smoking and Health. 2004. Accessed November 27, 2009.
<http://www.cdc.gov/tobacco/data_statistics/sgr/2004/index.htm>

3

U.S. Department of Health and Human Services. Reducing the Health Consequences of
Smoking: 25 Years of Progress. A Report of the Surgeon General. U.S. Department of
Health and Human Services, Public Health Service, Centers for Disease Control, Center
for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
DHHS Publication No. (CDC) 89-8411. 1989.

4

U.S. Department of Health and Human Services. Tobacco Use Among U.S.
Racial/Ethnic Minority Groups-African Americans, American Indians and Alaska
Natives, Asian Americans and Pacific Islanders, and Hispanics: A Report of the Surgeon
General. Atlanta, GA: U.S. Department of Health and Human Services, CDC. 1998.
Accessed December 5 2009.
<http://www.cdc.gov/tobacco/data_statistics/sgr/1998/index.htm>

5

Ockene IS, Miller NH. Cigarette Smoking, Cardiovascular Disease, and Stroke: A
Statement for Healthcare Professionals From the American Heart Association Journal of
American Health Association. 1997; 96(9): 3243–3247.

6

Carlsen E, Giwercman A, Keiding N, Skakkebæk NE. Evidence for decreasing quality
of semen during past 50 years. BMJ. 1992; 305: 609-13.

7

Swan SH, Elkin EP, Fenster L. The question of declining sperm density revisited: an
analysis of 101 studies published 1934-1996. Environ Health Perspect. 2000; 108: 961-6.

71

8

Kaufman RB, Spitz AM, Strauss LT, Morris L, Santelli, JS, Koonin LM, and Marks JS.
The decline in US teen pregnancy rates, 1990-1995. Pediatrics. 1998; 102: 1141-1147.

9

Pearce D, Cantisani G, Laihonon A. Changes in fertility and family sizes in Europe.
Popul. Trend. Spring. 1999: 33-40.
10

Mendiola J, Jorgensen N, Minguez-Alarcon L, Sarabia-Cos L, Lopez-Espin JJ, ViveroSalmeron G, Ruiz-Ruiz KJ, Fernandez MF, Olea N, Swan S, Torres-Cantero AM. Sperm
counts may have declined in young university students in Southern Spain. Andrology.
2013: 1-6.

11

Jorgensen N, Nordstrom UN, Jensen TK, Jensen MB, Almstrup K, Olesen IA, Juul A,
Andersson AM, Carlsen E, Petersen JH, Toppari J, Skakkebaek NE. BMJ Open. 2012
Jul 2; 2(4).

12

Rolland M, Le Moal J, Wagner V, Royere D, De Mouzon J. Decline in semen
concentration and morphology in a sample of 16 609 men close to general population
between 1989 and 2005 in France. Hum Reprod. 2013; 28(2): 462-70.

13

Lutz W. Fertility rates and future population trends: will Europe’s birth rate recover or
continue to decline? Int J Androl 2006 February; 29(1): 25-33.

14

United Nations. World Population Prospects: The 2006 Revision. New York, United
Nations 2007.

15

Bonde, JP, Ernst E, Jensen TK, Hjollund, NHI, Kolstad H, Henriksen TB, Scheike T,
Giwerman A, Olsen J, Skakkebaek NE. The relation between semen quality and fertility.
A population based study of 430 first-pregnancy planners. Lancet. 1998; 352: 11721177.

16

Connolly GN. Worldwide expansion of transnational tobacco industry. Journal of the
National Cancer Institute Monographs. 1992; 12: 29-35.

72

17

International Agency for Research on Cancer. IARC monographs on the evaluation of
the carcinogenic risk of chemicals to humans: Tobacco smokmg. IARC Mongraphs on
the Evaluation of the Carcinogenic Risk of Chemicals to Humans. 1986; No. 38: 65-74.

18

Ostergaard, K. Cadmiumkoncentrationen inyrevaev hos rygere og ikke-rygere.
Ugeskr. Lag. 1977; 139: 989-991.
19

Kier LD, Yamasak E, Ames BN. Detection of Mutagenic Activity in Cigarette Smoke
Condensates. Proc. Nat. Acad. Sci. USA. 1974; No.10: 4159-4163.

20

Kaufman DW, Helmrich SP, Rosenberg L, Meittinen OS, Shapiro S. Nicotine and
carbon monoxide content of cigarette smoke and the risk of myocardial infarction in
young men. N. Engl. J. Med. 1983; 309: 409-414.

21

Ravenkolt, RT. Radioactivity in cigarette smoke. New Engl. J. Med. 1982: 307-312.

22

Mattison, DR. The effects of smoking on fertility from gametogenesis to implantation.
Environmental Research. 1982; 26(2): 410-433.

23

Vogt, HJ, Heller WD, Borelli S. Sperm quality of healthy smokers, ex-smokers, and
never-smokers. Fertil. Steril. 1986; 45: 106-110.

24

National Research Council Subcommittee on Reproductive and Neurodevelopmental
Toxicology. Chapter 7: Biologic markers of human male reproductive health and
physiologic damage. In: Biological Markers in Reproductive Toxicology (eds N.
Grossblatt & L. R. Paulson), pp. 83-105. National Academy Press, Washington, DC.
1989.
25

Trummer, H., Habermann, H., Haas, J., Pummer, K. The impact of cigarette smoking
on human semen parameters and hormones. Human Reprod. 2002; 17: 1554–1559.

26

Kunzle R, Mueller MD, Hanggi W, Birkhauser MH, Drescher H, Bersinger NA.
Semen quality of male smokers and nonsmokers in infertile couples. Fertility and
Sterility. 2003. 79(2): 288-291.
73

27

Sepaniak S, Forges T, Gerard H, Foliguet B, Bene MC, Monnier-Barbarino P. The
influence of cigarette smoking on human sperm quality and DNA fragmentation.
Toxicology. 2006; 223: 54-60.

28

Chia SE, Lim SA, Tay SK, Lim ST. Factors associated with male infertility: a casecontrol study of 218 infertile and 240 fertile men. BJOG. 2000; 107(1): 55-61.

29

Zhang JP, Meng QY, Zhang LJ, Mao YL, Sun ZX. Effects of smoking on semen
quality of infertile men in Shandog, China. Asian J Anrol. 2000; Jun; 2: 143-146.

30

Sofikitis, N Miyagawa I, Dimitriadis, D et al. Effects of smoking on testicular
function, semen quality and sperm fertilizing capacity. J. Urol. 1995; 154: 1030-1034.

31

Chen HW, Kuo CT. Cotinine Characterization and Quality Effect of Sperm for
Smoking and Nonsmoking Students. Bull Environ Contam Toxicol. 2007; 79: 11-14.

32

Wong WY, Thomas CM, Merkus HM, Zielhuis GA, Doesburg WH, and SteegersTheunissen RP. Cigarette Smoking and The Risk of Male Factor Subfertility: Minor
Association Between Cotinine in Seminal Plasma and Semen Morphology. Fertil Steril.
2000; 74(5): 930-5.

33

Zenzes MT, Bielecki R, and Reed TE. Detection of benzo(a)pyreene Diol EpozideDNA Adducts in Sperm of Men Exposed to Cigarette Smoke. Fertil Steril. 1999; 72(2):
330-5.

34

Sofikitis N, Takenaka M, Kanakas N, Papadopoulos H, Yamamoto Y, Drakakis P, and
Miyagawa I. Effects of Cotinine on Sperm Motility, Membrane Function, and Fertilizing
Capacity In Vitro. Urol Res. 2000; 28(6): 370-5.

35

Usmani ZC, Craig P, Shipton D,and Tappin D. Comparison of CO breath testing and
women's self-reporting of smoking behaviour for identifying smoking during pregnancy.
Subst Abuse Treat Prev Policy. 2008; 3:4.

74

36

Uematsu T, Mizuno A, Negashima S, et al. The axial distribution of nicotine content
along hair shafts as an indicator of changes in smoking behavior. Evaluation in a
smoking cessation programme with or without the aid of nicotine chewing gum. Br J
Clin Pharmacol. 1995; 39: 655-699.
37

WHO. World Health Organization Laboratory Manual for the Examination of Human
Semen and Semen-Cervical Mucus Interaction. 3rd ed. Cambridge: Cambridge
University Press, 1992.

38

WHO. World Health Organization Laboratory Manual for the Examination of Human
Semen and Semen-Cervical Mucus Interaction. 4th ed. New York: Cambridge University
Press, 1999.

39

Kruger TF, Menkveld R, Stander FS, Lombard CJ, Van der Merwe JP, van Zyl JA, et
al. Sperm morphologic features as a prognostic factor in vitro fertilization. Fertility and
Sterility. 1986; 46: 1118-23.
40

Munkelwitz R, and Gilbert BR. Are boxer shorts really better? A critical analysis of
the role of underwear type in male subfertility. J Urol. 1998; 160(4): 1329-33.

41

Gebreegziabher Y, Marcos E, McKinon W, and Rogers G. Sperm Characteristics of
Endurance Trained Cyclists. Int J Sports Med. 2004; 25(4): 247-251.
42

De Souza MJ, and Miller BE. The effect of endurance training on reproductive
function in male runner. A ‘volume threshold’ hypothesis. Sport Medicine (Auckland,
N.Z.) 1997; 23(6): 357-374.

43

Wise LA, Cramer DW, Hornstein MD, Ashby RK, and Missmer SA. Physical activity
and semen quality among men attending an infertility clinic. Fertility and Sterility. 2011;
95(3): 1025-1030.

44

Sanger WG, et al. Fit of underwear and male spermatogenesis: a pilot investigation.
Reprod Toxicol. 1990; 4(3): 229-32.

75

45

Parazzini F, Marchini M, Luchini L, Tozzi L, Mezzopane R, and Fedele L. Tight
underpants and trousers and risk of dyspermia. Int J Androl. 1995; 18(3): 137-40.

76