LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 chIRC/DanDuopss-sz

EFECTS OF ORGANOCHLORINE COMPOUNDS AND HEAVY METALS ON
MALE REPRODUCTIVE HEALTH

By

Julia Jennifer Wirth

A THESIS

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

MASTER OF SCIENCE
Department of Epidemiology

2001

ABSTRACT
EFECTS OF ORGANOCHLORINE COMPOUNDS AND HEAVY METALS ON
MALE REPRODUCTIVE HEALTH
By

Julia Jennifer Wirth

Ten to 17% of American couples currently seek medical help for infertility; half of these
problems have a male cause usually of unknown etiology. Wildlife, experimental animal
and human epidemiological studies indicate that environmental contaminants,
especiallyorganochlorine compounds (OCs) such as polychlorinated biphenyls (PCBs),
TCDD (dioxin) and DDT/DDE and heavy metals have adverse effects on male
reproduction. To investigate the relationship between human male reproductive health
problems, specifically semen quality and reproductive hormone levels (FSH, LH,
testosterone and inhibin B), environmental contaminants and polymorphisms in genes
involved in contaminant and sex steroid metabolism (P450), we propose to conduct a
cohort study recruiting men based on exposure to Great Lakes sport-caught fish meals
(none, 1 to 1 1 and greater than 12 meals) from infertility clinics in three areas of
Michigan. Information on Great Lakes sport-caught fish consumption, as well as on other
risk factors, and fertility will be obtained from a self-administered questionnaire.
Secondary exposures (contaminants present in Great Lakes fish), including PCBs, dioxin,
DDT and heavy metals, and P450 polymorphisms, will be measured from a blood
sample. This study provides the opportunity to investigate the association between
measures of male fertility, known reproductive toxicants found in fish, and gene-

environment interactions affecting semen parameters and reproductive hormone levels.

To My Friends

Without whom this would not have been possible

iii

ACKNOWLEDGEMENTS

I would like to thank my thesis advisor, Dr. Nigel Paneth, for his guidance,
encouragement and faith that an idea could be transformed into a grant proposal, and for
exemplifying a true scholar and mentor. Thank you for providing me the opportunities

for growth and change.

Many thanks to Dr. Claudia Holzman and Dr. Larry Fischer, my committee members, for

your their patience, advice and invaluable input.

Heart felt gratitude to the students and staff of the Department of Epidemiology.

Special thanks to Mary Fran Steele, Jan Pylar, Mel Kennicott, Julia Freije, Jeff Wirth,

Andrew Mullard, and Teri Sexton for helping me to see the light.

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................ vii
LIST OF FIGURES ....................................................................... viii
LIST OF ABBREVIATIONS ............................................................. ix
A. SPECIFIC AIMS ........................................................................ l
B. BACKGROUND AND SIGNIFICANCE ........................................... 2
1. Changes in male reproductive health .................................... 2
a. Changes in semen quality ................................................... 2

l) Carlsen et al.’s study and the decline in semen quality... . . . .....3
2) Relationship between semen quality and fertility. . . . . . . . .....24
b. Changes in the incidence of abnormalities of the

male genital tract .......................................................... 28

c. Changes in the sex ratio ................................................... 37

2. Exposure to OCs and mercury ............................................ 38
a. OCs and mercury in the environment ................................... 38
b. OCs and mercury levels in fish ........................................... 49

c. OCs and mercury levels in human ...................................... 5]

3. Effects of OCs on male reproductive outcomes ........................ 55
a. Observations on wildlife and their exposures ......................... 55
b. Effects on spermatogenesis ............................................... 57
c. Effects on the sex ratio ................................................... 68
d. Comparison of animal and human OC doses .......................... 75

4. Mechanisms of OCs effects on male reproductive parameters... ..78
a. Role of genetic polymorphisms .......................................... 78

b. Role of direct toxicity and reactive oxygen species (ROS) ........... 87
5. Effects of mercury on male reproduction ............................... 89
C. PRELIMINARY STUDIES ............................................................ 90
1. Previous result from the FFHP ................................................ 90
2. Pilot Study: Michigan Men’s Study ......................................... 92
3. Research team ................................................................... 93
D. RESEARCH DESIGN AND METHODS ........................................ 95
1. Overall design .................................................................. 95
a. Study areas ................................................................. 97

b Recruitment ................................................................ 98

c Substudy .................................................................. 100

d. Questionnaires ........................................................... 100

e. Training of recruiters ..................................................... 101

2. Data collection .................................................................. 101

a. Measurement of consumption of Great Lakes fish .................. 101
b. Other risk factors ......................................................... 102
c. Laboratory tests .......................................................... 103
3. Data analysis and interpretation ............................................. 1 10
a. Data analysis .............................................................. 110
b. Sample size and power calculation ..................................... 112
c. Project timetable ......................................................... 1 14
4. Strengths and weaknesses ................................................... l 15
E. HUMAN SUBJECTS ............................................................. l 17
F. REFERENCES ..................................................................... 1 18
APPENDIX A
WHO GUIDELINES FOR SEMEN ANALYSIS PARAMETERS ............ 150
APPENDIX B
SUMMARY OF CHEMICALS QUANTIFIED IN FISH TISSUE .............. 151

TRIGGER LEVELS USED BY MDCH TO ESTABLISH
CONSUMPTION WARNINGS ...................................................... 152

vi

LIST OF TABLES

Table 1 Estimated recruitment of most exposed men by clinic ................. 99
Table 2 Sample size requirements ................................................. 113
Table 3 Power estimates ............................................................. 1 14

vii

LIST OF FIGURES

Figure 1 Structure of DDT, PCBs, dioxin, and dibenzofurnans .................. 33
Figure 2 Positions of chlorines on PCBs ............................................ 34
Figure 3 Model for dioxin activation of genes via AhR binding ................. 79

viii

LIST OF ABBREVIATIONS

CI ............................................................................... Confidence Interval

OCs ................................................................... Organochlorine Compounds
OR ......................................................................................... Odds Ratio
FFHP ............................................................ Fisheaters Family Health Project
FSH ................................................................. Follicle Stimulating Hormone
LH ............................................................................ Luteinizing Hormone
PCBs ................................................................... Polychlorinated Biphenyls

TCDD ............................................... 2,3,7,8-Tetrachlorodibenzodioxin (dioxin)
TE ....................................................................................... Testesterone

ix

 

{Mir I... {limb-rill. . you"

A. SPECIFIC AIMS

To better understand the relationship between male infertility and environmental
contaminants, we pr0pose to conduct a retrospective cohort study recruiting men based
on exposure from 3 infertility clinics in Michigan. Specifically, the aims of this proposal
are to:

1. Recruit men attending infertility clinics in 3 areas of Michigan abutting Great

Lakes, based on three levels of sport-caught Great Lakes fish consumption in the

previous yearz378 men who consumed 12 or more meals (most-exposed); 378 men

who consumed less than 12 but at least one meal (moderately-exposed); and 378

those who consumed no meals (unexposed); matched within 5 years of age, within 2

months of enrolment and by clinic;

2. For subjects at all exposure levels:

a. Obtain blood samples for analyses of reproductive hormones (TE, FSH, LH,
inhibin B and estradiol), environmental contaminants (organochlorines, mercury
and polymorphisms in genes involved in contaminant and steroid
metabolism (cytochrome P450, glutathione S-transferase F{GST});

b. Assess sperm number, motility, and morphology and semen quality;

c. Administer a questionnaire on reproductive health, fish-eating habits, occupation

and general demographics;
3. Collect semen samples from a subsample of men with high (upper quartile) and low

(lower quartile) levels of PCB exposure, as determined in serum samples, to perform

more specific and sensitive tests of sperm function;

4. Test the hypotheses that compared to unexposed subjects, moderately and most
exposed subjects will have increased risks of altered sperm function parameters and
increased risks of having abnormal levels of reproductive hormones.

5. Test the hypothesis that polymorphisms in cytochrome P450 and GST genes will
interact with environmental contaminants (organochlorines, mercury) to
produce an increased risk of alterations in sperm function and/or reproductive

hormone levels in moderately and most exposed subjects.

B. BACKGROUND AND SIGNIFICANCE

1. Changes in male reproductive health
Reproductive problems are a concern for a growing number of American couples with
between 10 and 17% seeking medical help. In 1987 about $1 billion dollars were spent on
treatments for infertility ( l ). In about half of the cases, the problem is male and up to
80% of these cases have an unknown etiology (2, 3). Concern has focused on the reported
changes in markers of male reproductive health, including decreasing sperm densities,
increasing incidence of disorders of male reproductive development, including

cryptochidism, hypospadias and testicular cancer, and decreasing sex ratios.

a. Changes in semen quality
Reports have been published both supporting and refuting a world-wide decline in sperm
density over the last 50 years. A meta-analysis published in 1992 by Carlsen et a1.(4),
concluded that sperm densities have declined internationally by at least 50% in the past

three decades. The study was criticized on methodological grounds (5) (6) and touched

off heated debate on the validity of the findings. The debate highlighted three
controversial issues regarding the possible decline in male reproductive health: the
conclusions of Carlsen et al.’s study in particular and the decline in semen quality in
general; the association between semen quality and fertility; and the increase in
congenital abnormalities of the male genital tract.

1) Carlsen et al.’s study and the decline in semen quality

a) Summary of Carlsen et al’s study

In 1992 Carlsen et a1.(4) in a meta-analysis of 61 studies, concluded that sperm density
had declined significantly between 1940 and 1990. Using linear regression analysis of
the data weighted for the number of men in the individual studies, they found that the
mean sperm count decreased from 1 13 x 106/m1 to 66 x 106/ml. The estimated regression
coefficient was —0.934x106/m1 per year (SE: 0.157, p<0.0001). The mean count rather
than the median was used since the median could only be derived from 19 studies.
Additionally, in 46 of the 61 studies in which the seminal volume was reported, linear
regression showed a significant decrease during the same time period from 3.40 ml to
2.75 ml with an estimated linear regression coefficient of —0.0130 m1/year (SE: 0.0057,
p<0.027). The meta—analysis did not include sperm motility or sperm morphology
because their evaluation was considered to be rather subjective and values for them were

not included in older publications.

The 61 studies included 14,947 men; 39 publications (8428 men) included men with
proven fertility only, while 22 publications reported on men unselected with respect to

fertility status. These latter men were considered to represent the “normal male

 

 

 

 

 

 

dft

im

population” (their quotation marks). A separate analysis of mean sperm density including
only studies with men with proven fertility had a regression coefficient of —1.062 x106/ml
per year (SE=0.185, p<0.0001). Information on age was provided in 42 of the studies and
ranged from 17 to 64 years, with a mean of 30.8 years. A minimum period of sexual
abstinence of three days was requested in 32 publications. The authors state, however,
that the recorded data did not allow them to analyze the effects of the period of
abstinence on semen parameters. While 21 countries from North and South America,
Europe, Scandinavia, Africa, India and Hong Kong were represented, 28 of the reports
came from the USA.
b) Criticism of Carlsen’s et a]. study and the decline in semen
quality
Carlsen et al.’s report was criticized on methodological issues, including the non-
comparability of the studies and the appropriateness of the methods used in their
statistical analysis.
1)) Non-comparability of the studies
The criticism concerning the non-comparability of the studies focused on the
performance of semen analyses for the individual studies in different laboratories, and the
use different study populations.
a)) Use of different laboratories for semen analysis
The major concern is the possible introduction of variation when the semen analyses that
are being compared are performed in different laboratories. As a consequence, spurious
differences in semen parameters that arise due to different laboratory methods may be

interpreted as true differences in sperm density over time and across studies.

lll'

A semen analysis involves few or many measures of sperm number, morphology and
function as well as semen quality (pH, consistency, color). The parameters most
frequently found in publications are the quantitative measures including sperm count
(reported in millions), semen volume (ml), sperm density or concentration (millions per
ml), and total sperm count per ejaculate (millions). Sperm counts should be presented as
the median count, since the distribution is usually skewed and the median value is a better
measure of central tendency. The sperm count itself is the parameter least susceptible to
error (7). Sperm motility, morphology and viability are qualitative measures (8) (9), may
be reported in different ways (10) and are thus more likely to involve technical error.
Sperm count, sperm motility and sperm morphology are also subject to variation (1 1)
(12). A within-subject variability of 50 to 75% of the total variability in sperm
concentration, semen volume and motility measurements has been reported (13). Another
study conducted as part of a longitudinal study of human semen characteristics of
unexposed workers measured sperm motility variables Using computer- assisted sperm
analysis (14). Motility analyses were conducted on monthly samples from 46 men for 9
months. The variability within a sample, between samples from the same individual
(between monthly samples), and between individuals were calculated using a nested
analysis of variance. For all sperm motility measurements, at least 90% of the variation
was observed between microscope fields within a semen sample. For all sperm motility
variables, variation between subjects was the smallest percentage, ranging from 1.3% to
4.0%. When sample means were used in the nested analysis of variation, at least 75% of

the variation was observed between samples from the same individual.

 

L1

.1?

0?

all

TCI
ll'li‘t
\"lfli

Van

Variation in the parameters of an individual’s semen analysis is due mainly to
measurement errors (15). Counting or statistical errors leading to lack of precision are
associated with counting limited numbers of sperm. The exact number of sperm in a
given volume follows a Poisson distribution where the variance is equal to the number
counted, making the variation in count highly dependent on the number of sperm
counted. Other random measurement errors include inadequate mixing of the semen
sample, technician stress, poor technique and instrument variation (16). In fact, repeated
analysis by the same technician using the same procedure will yield greater variation in a
semen analysis than the counting error alone (15). In Carlsen et al.’s meta-analysis, errors
of this sort would be expected to be randomly distributed amongst the studies and would

affect the precision of the measurements, but not the validity.

Temporary life events can affect the semen analysis parameters. Febrile illnesses,
infections, unusual stress, and seasons (17) (18) (reviewed in Clark (19) all can cause
significant variations in sperm counts and thus affect the precision of the measured
variables (13) (20). Increasing scrotal temperature from the optimum of 35°C can impair
spermatogenesis (21) (reviewed in (22)). Different types of clothing, by their insulating
effects, can raise scrotal temperature (23) and alter sperm function. A prospective
randomized trial was conducted with healthy men of proven fertility between the ages of
25 and 50 years testing the effect of wearing tight-fitting versus loose-fitting underwear,
24 hours a day for 6 months, and sperm density and motility (24). After 6 months, the

men switched protocols, wearing the other type of underwear for another 6 months.

m

b:

la

\ll

Du

8pc

by

abs?

Using a matched-pairs signed rank test, they found that sperm density was significantly
reduced from 89.5 x106/ml to 46.0 x106/ml (p<0.05) and the concentration of motile
sperm was decreased from 53.1x106/ml to 17.4x106/ml (p<0.005). The number of men
participating, however, was low: 20 volunteers started the study and only nine completed
it. Another study, also testing the effects of wearing tight versus loose—fitting underwear
and using the same alternating protocol, but with even fewer men (n=2) also found a
decline in sperm density and motility with duration of exposure (25). Several other
studies have found similar results, although never with a sufficiently large study group

(26) (27).

Seasonal changes have also been reported to affect sperm density, semen volume and
motility by some (28) but not all (13) investigators. The seasonality of sperm counts may
be more pronounced for studies taking place in areas with large yearly temperature
variations (13). Information on these events was not obtained by Carlsen et a1., but their
variation would be expected to be randomly distributed amongst the subjects in the
studies, especially if the recruitment phase of the study lasted at least a year to include all

seasons. Thus, the validity should not be affected by these variables.

Duration of abstinence prior to semen collection is another life event that can affect
sperm parameters (29). Values for the quantitative parameters have been reported to vary
by threefold between 1 and 7 days of abstinence (16), with longer periods of sexual
abstinence resulting in higher sperm counts and lower percentages of motility (30) (31).

Sperm counts have been estimated to increase by 5-13 x 106 per ml per additional day of

 

.ll‘xlll‘
agani
HIMC
comrl
reducc
Wurltl
\ltllltltl:
length

nundar

Not all 5
mducho
lulut‘s fr
aficrone
hmhm
decreaset
make II II
POPUIJIK)
dfldloudg
fioninnér
m44d®(
ElIlluated

“Ekm;

 

abstinence (32). If the duration of abstinence is not specified in a study, which is rare, it
again should be randomly distributed among the subjects and only affect precision.
However, if different studies recommended different durations and did not record and
control for the differences in the analysis, it could affect the validity. In an effort to
reduce variation in duration of abstinence as well as in other semen parameters, the
World Health Organization (WHO), beginning in 1980, published a manual on
standardized methods for evaluating semen analysis parameters, and recommended a
length of abstinence between 2 and 7 days. Prior to that time there were no written

standards.

Not all studies have found significant effects of length of abstinence, however. A
reduction of only approximately 8% of control mean sperm concentration values (mean
values from samples collected after abstinence of 3 to 5 days prior to entering the study)
after one day of abstinence was found for 12 healthy men, aged 18 to 25 years (29).
Total sperm count and semen volume, but not sperm motility or viability, were similarly
decreased. The small sample size and restricted age range of the subjects in this study
make it impossible to generalize the very small decrease in sperm density to the general
population. Another study also found little association between duration of abstinence
and total sperm count (33). A decrease in duration of abstinence was observed in males
from infertile couples in Sweden from 7.5 days in 1956 and 1966 to 5.0 days in 1976 and
to 4.4 days in 1986 (33). The total number of men evaluated was 785, with at least 141
evaluated per time period. Donors from 1956 were men whose partners were considered

to be fertile, while those from 1986 donated semen prior to assignment of fertility status.

 

517C
and
abs

dcc

The;
33pa
da}5_

uudjc
andrnl
Additl
’nauur
[he Per

andlhc

These investigators found a significant decease in total sperm count from 467 x 106 in
1956 to 305 x 106 in 1986 (p<0.0001). as well as a significant change in percentage of
sperm with normal morphology (53% versus 37%, p<0.0001). When they restricted their
analysis to men reporting 3 to 5 days of abstinence versus no restriction on the period of
abstinence, the total sperm count recorded at each year (1956, 1966 and 1976) did not
decrease significantly, and the difference between counts from 1956 and 1986 remained
significant (p<0.05). The change in morphology, which is not affected by length of
abstinence (17) (34) (29), was unchanged. While it is clear that sperm counts and semen
volume vary according to duration of abstinence, it is unclear if the reduction is always
significant. A period of abstinence from one to seven days may have little significant
effect on sperm counts. However, if the period of abstinence is not specified in the study

design and the actual length not recorded, it can be a confounder for sperm concentration.

The period of sexual abstinence was recorded by Carlsen et al., but was available for only
32 papers. For these 32 studies, the prescribed length of abstinence was at least three
days. They did not, however, do a separate analysis of these studies. For the remaining
studies, Carlsen et al. note that while the period of abstinence was not mentioned,
andrologists have recommended a period of three days for the past 50 years.

Additionally, to the best of their knowledge, Carlsen et al. state that no change in
masturbation or coital frequencies had occurred since the 1930’s, implying that whatever
the periods of sexual abstinence for these studies were, they haven’t changed over time

and therefore should not confound their conclusions.

Twenty—seven of the studies in the meta-analysis were published before 1980 and would
not have had access to the WHO standards on period of sexual abstinence. There is still
no evidence that studies published after 1980 adhered to the WHO standards, or if they
just recommended a specific period of abstinence or if they recorded the actual period.
Carlsen et al. argue that the unwritten standard since the 1930’s was three to five days. A
quick inspection by myself of 10 publications included in the meta-analysis with semen
analyses performed from 194lto 1993 revealed that all recommended at least three days,
three to five days or three to ten days of abstinence, but none recorded the actual length
of abstinence or controlled for it in their analyses. However, the semen analyses were
performed in the same laboratory with the same standards over the indicated time span.
It thus seems likely that the unwritten period of abstinence of three to at least five days
was requested in many of the studies. Unless there is reason to think that the rate of
compliance has changed over time, the length of abstinence has probably changed little
over time and the various lengths should be randomly distributed within and between the

studies.

Several studies have recorded changes in coital rates over time and, it is argued (16) these
changes may have affected sperm densities. As mentioned above, Bendvold (33) reported
a change in abstinence from 1965 to 1986 but found it had little effect on sperm counts or
semen volume. A 21% age-standardized increase in marital coital rates in the US from
1965 to 1970 (7.0% in 1965 to 8.2% in 1970) was reported (35). The study compared
responses of currently married women under age 45 years participating in the National

Fertility Studies to a question on the frequency of intercourse in the last four weeks. The

10

 

€‘\

be

Gilt

author acknowledged a tendency to report stereotypical frequencies i.e. eight times per
month. Additionally when the sample of 440 women were re-interviewed four to seven
months later in 1965, 68% of the women gave responses with a range of plus or minus
two days, which is greater than the reported difference in the frequencies from 1965 to
1970. It is also doubtful that the male partners of these women are representative of
males in the general population or the population of men attending fertility clinics. James
presented data from Trussell and Westoff (1980) and Abma (1993) showing a 22%
increase in coital rates between 1965 and 1975 in married US couples and a subsequent
decrease of 27% between 1975 and 1988 (36). The rise and fall in coital rates coincided
with a rise and fall in the sex ratio, but no attempt was made to establish causality, so
these two set of events remain a coincidence. Thus, the data supporting the effects of
changing lengths of abstinence over time provided by these reports are weak or non
existence, and these are the reports most frequently cited to illustrate the association
between period of abstinence and changing sperm count values (16). Thus, the data on

changing coital rates does not appear to support an effect on sperm density.

Another possible source of systematic error in sperm density measurements raised by
Carlsen et al.’s critics is the differing counting techniques and instruments Used by
different laboratories included in the meta-analysis. Changes in laboratory procedures,
such as the type of counting chamber used and the technical training of the individual
performing the analysis, can result in coefficients of variation between labs of 23-87%
(1 1), although others have found no effect of different counting chambers or

modifications to laboratory procedures (13). Laboratories may also differ in the analytic

ll

 

\d

ral

Iii;

attc
[hm
attri
mor]

hi)“:

Curls
Carly
”1&3th
hate I:

dk‘knl’fil

methods used to evaluate sperm function (37) (38), which can also introduce systematic
bias. For example, the determination of sperm morphology has changed since the
introduction of “strict” criteria by Kruger et al. in 1986 (8). With this classification, a
“normal” sperm must have a normal sperm head, neck, midpiece and tail. Any borderline
form is considered abnormal. using this classification, the predictive value of semen
samples containing at least 14% sperm with normal morphology for success in in vitro
fertilization and pregnancy is good (81% in vitro fertilization rate and 22% pregnancy
rate per embryo)(8). If comparisons are made between sperm morphology determined by
the older criteria and the “strict” criteria of Kruger, differences due solely to the use of
different criteria may be found. For example, a comparison of sperm morphology in men
attending an infertility clinic in 1974- 1984 with that of men attending the same clinic in
1991, revealed a decrease in the percent of morphologically normal sperm (20). The men
attending the clinic in 1991 were expected to have a higher percent of normal sperm than
those attending from 1974- 1983 since their overall fertility was higher. The authors
attributed the difference to the change in the criteria used to determine normal
morphology. The possibility that environmental exposures contributed to the change,

however, was not considered.

Carlsen et a1., cite WHO as recommending the USe of different counting chambers (39).
Carlsen et al. also excluded studies that USed computer assisted or flow cytometric
methods for evaluating sperm counts as these methods were not available until 1980 and
have lead to inaccurate counts for samples with low sperm densities (40). They

acknowledge the apparent imprecision of sperm counting, but find no reason to believe

12

 

.Ax

C‘s.

\lal
CXdl

‘1

“Cr.

rang:
\Pcrn
head.
Vanat
Ihcla
ln(:ar
‘lgnifi

d€5Plh

that the test is subject to secular trends. In support of this, they mention that
hematologists who have USed the same counting chambers for the past 50 years, have

not reported a similar secular trend in blood cell counts (4).

As mentioned earlier, in an effort to reduce the variation in semen analyses performed in
different laboratories, WHO publishes guidelines for evaluating semen. The WHO
Manual for the Evaluation of Human Semen provides standards for collection and
examination of human semen, sperm preparation techniques, quality control measures as
well as references values for semen parameters. (The current WHO normal values for
semen parameters are listed in Appendix A). Despite this attempt at standardization, a
large multi-center trial on external quality control between eight laboratories in Germany
in 1990 found large coefficients of variation for sperm count, motility and morphology
between different laboratories (1 1). Coefficients of variation (CV) for sperm counts
ranged from 23% to 73% for samples with high and low concentrations, respectively. For
sperm morphology, the range of CVs for the different sperm parts was 25% for normal
heads and 87% for abnormal midpieces, and the CV for motility was 21%. Some of the
variation in sperm counts, which were made on formalin-fixed specimens and mailed to
the laboratories, may have been due to the shipping conditions. Since most of the studies
in Carlsen et al.’s meta-analysis used different laboratories, it is likely that there was
significant variation in methodologies and in sperm density. This would seem to be true

despite publication of the WHO guidelines.

 

 

C0

fen

bib]

C(llll

infer

Were

pre-r

10 be

men.

lp<0

10 be

b)) Use of different study populations
Another major criticism of the Carlsen et al.’s study was the inclusion of studies in the
meta-analysis with populations too diverse to be compared so that no meaningful
conclusion could be made. These population differences were fertility status and the
geographical and racial composition of the participants in the different studies.

1)) Fertility status

The meta-analysis included data on men unselected with respect to fertility as well as on
men with proven fertility. Their critics (6) (41) (42) (16) argue that men with proven
fertility, which is defined as having fathered at least one child, and men with unknown
fertility will differ with respect to semen parameters, including sperm count (10). The
basis for this argument is that men who have fathered a child will have higher sperm
counts than men unselected for fertility status, which presumably includes men who are
infertile as well as fertile men. In fact, the men from the studies included in this group
were sperm donors (43), men reporting for vasectomies (43) (18) and men coming in for
pre-marital check-ups (43). In the study by Wang et a1., 95% of the men were considered
to be fertile. Carlsen et al. state that when they analyzed studies containing only fertile
men, the regression coefficient for men sperm concentration was —8.52 x 106/ ml per year
(p<0.0001), which is very close to the values for all of the studies. So there does not seem

to be any bias regarding fertility status.

It is also possible, however, that as the knowledge and treatment of infertility problems

has increased, men with lower sperm counts have become fathers. Thus these men would

be included in the category of fertile men in the more recent studies, but would have been

14

 

l 0.

Hr:

mill

considered in- or sub-fertile in the earlier studies and possibly not included in some
studies. Consequently, the definition of normal sperm count has changed over time (6)
Olsen 1994). In the earlier studies from the 1940’s (44), a sperm concentration of 60 x
106/m1 was considered normal, while currently a concentration of 20 x 106/ml is
considered normal (15). If this change led to exclusion of men in the earlier studies with
sperm counts less than 60 x 106/ml (that would be included in the later studies), this
would artificially increase the mean of the earlier sperm counts. In support of this
possibility Bromwich et a1. (6) published a report criticizing Carlsen et al. on their
statistical analysis. Bromwich et al. used various mathematical models of the probability
distribution of mean sperm concentrations and then provided their own data from men
attending an in vitro fertilization clinic to show that the distribution is heavily skewed
towards lower counts. They argue that the data best fit a logarithmic distribution and that
truncation of such a distribution at 60 x 106 sperm/ml alone would account for the drop in

sperm count observed by Carlsen et a1.

However, as pointed out by Keiding et al. (45), the studies from the 19405 (17) include
many sperm counts lower than 60 x 106 sperm/ml, suggesting that the data was not
truncated. Of interest, the sperm density distributions of these earlier studies (reviewed in

(17) are actually skewed toward the higher concentrations.

2)) Comparison of studies with wide variation in geographic

distribution

15

 

 

pu

U 1‘

ant

To

Go}
gar:
rent
01' m

197i

Another criticism of the populations compared in Carlsen et al.’s study involves the
geographic distribution of the reviewed publications. The meta-analysis included
publications from all over the world although almost half were from the USA. Inclusion
of such diversity is sure to increase the variability of the counts due to genetic, life-style
and environmental exposure differences of the populations as well as differences in the
technical analysis of the sperm counts. On the other hand, there is no other way, given the

data, to assess global changes.

To determine if geographical variations influenced Carlsen et al.’s findings, Fisch and
Goluboff (46) re-analyzed the data using only those studies with at least 100 men. This
gave them 20 publications, which still contained 91% of all the men in Carlsen et al.’s
review. They noted that all 5 papers before 1970, which contained the high (100 x 106/ml
or more) sperm counts, were from the USA, and 4 of the 5 were from New York. After
1970, when the counts were lower, only 3 out of 15 were from the USA and only one was
from New York. Of the 7 papers reporting the lowest values, 5 originated from
developing countries. Fisch and Goluboff concluded that, given the differences in
methodologies and in patient selection, geographic variations in sperm count need to be
considered when evaluating data from different locations. There is a problem with this
interpretation. All but 2 (n: 100 men) of the 13 studies in Carlsen’s review between 1938
and 1969 are from the USA (n=1680 men), so there is no other country with close to the
same number of studies or men to serve as a comparison to the USA data. All but two of
USA studies report counts above 100 x 106/ml. Thus there is no way to know if the high

sperm counts reported by these early USA publications are due to the time frame of

16

 

«’1

I“

the x;-
ehnir
TVCR'

PUDllr
urca.‘

fiach.

semen collection or to the location. Of note, even though they represent only 50 men, the
two studies not from the US reported counts of 103.2 x 106/ml (Peru, 50 men) and 87.5 x

106/ml (Denmark, 50 men) (in Carlsen et al. (4) ), both of which are in the high range.

The issue of location has been addressed in several publications looking at data collected
within countries. An early report by McLeod and Wang (7) compared studies from the
USA from early times (1938 to 1951 in New York) and the “modern” era from 1970 to
1977 in New York, Iowa and Houston. Using the mean, median or the frequency
distribution of sperm counts, they agreed that a depression of spermatogenesis had
occurred since 1951. However, when they used their own data from 1951 on infertile
men and compared it to data they collected from 1966 to 1977 on infertile populations,
they found no decrease. They suggest that the use of the same experimental conditions in
the same laboratory and essentially the same type of populations (infertile New Yorkers)
eliminated the differences seen in the other studies. One hopes, however, that infertile
New Yorkers are not representative of the USA males. On the other hand, Leto et a1. (47)
published a study on 275 semen donors from 1973 to 1980 from the Washington, DC
area. They noted that the average sperm concentration, which was a mean of 12 replicates

each, fell from 120 x 106 /ml in 1973 to almost 90 x 106 /ml in 1980.

Other investigators however have found an increase in sperm densities in studies within
the USA. Fisch er al. (48) compared mean sperm counts, sperm motility and semen
volume in three USA sperm banks (Roseville, MN, Los Angeles, CA and New York,

NY), from 1970 to 1994. They reported that sperm counts differed significantly (p<0.01)

l7

 

Luz.

bet

in"

“10'
slim
sper
foun
111.1}

agret
1\)lk
lktrk

lime.

I\nni
wt“ CI
Unneu
SeVCru
”mm.

anall'gj

between locations with a high in New York of 131.5 x 106/ml (mean) and a low of 72.7 x
106/ml in California. Using linear regression of sperm concentrations for the total
population, they found a significant increase from a mean of 77 x 106/ml in 1970 to 89 x
106/ml in 1994 (p<0.04). The also found an increase in sperm counts within the 3 centers,
which was significant for New York and Minnesota, and in men with proven or unproven
fertility status, both of which were significant. Finally in a multiple regression model
with age and duration of abstinence as potential modifiers, sperm concentration still
showed a significant increase (p<0.03). No data is presented on the distribution of the
sperm counts in this paper; however, if it is skewed towards lower counts, as has been
found in other studies (7) (6), the mean would be higher than the median and perhaps
may have biased their data towards higher values. Nevertheless, this study is in
agreement with McLeod and Wang (7) on the relatively high sperm counts from New
York and adds support to the suggestion that the sperm densities from the early New
York reports may have biased Carlsen et al’s analysis by virtue of location rather than
time. It also supports the notion of geographical variations in sperm densities in the USA

(7).

Another study aimed at addressing the effect of geographical location on sperm counts
was conducted in Seattle (49). Sperm counts from 510 healthy volunteers who served as
untreated controls for an intervention study from 1972 to 1993 were analyzed. Since
several semen samples were collected from each participant with a median number of 6
samples, the mean concentration was used in the analysis. Uniform methods of semen

analysis including duration of abstinence were followed. No change was observed in the

18

 

(l?

1
R

(H

Rcr
Eur
dctc‘
men
and
i‘crtil
31am
and r
"16 m
the 12
p0DUI.

1 07ml

geometric mean during this time period. In 1999, Saidi et a1. (50) published a review of
29 USA studies from 1938 to 1996, in which semen analyses were performed on 9,612
fertile or presumably fertile men. Mean sperm count was determined by geographic
region. Mean sperm concentrations from New York were significantly higher than other
USA cities (98.6 versus 71.6 x 106/ml, p=0.0006), and no significant change occurred
over time for any city. However, an analysis performed without separating by location
indicated a decline (p=0.047). Thus several studies from the USA on healthy men from
the same geographic areas using standardized methods for semen analysis as well as a
review found no change in sperm densities. Significant differences, however, were found

between different locations.

Reports from Europe, on the other hand, have found divergent results. Two studies from
Europe have also found a decrease in sperm count. In a study specifically designed to
determine if fertility in Danish men has declined from 1952 to 1972, semen analyses of
men from the Copenhagen area attending a single Sperm Analysis Laboratory in 1952
and in 1972 were compared (51). The men at both times were examined because of a
fertility problem. In 1952, 6.2% of the men had azoospermia while 3.9% of those men
examined in 1972 were azoospermic. The laboratory had one supervisor for the interval
and thus provided consistent methodology and technical skill. Additionally, the ages of
the men at both times were very close (3 year difference) but the social classification of
the 1972 men was higher, and both classifications were higher than that of the general
population. There was a significant decrease in the median sperm count from 73.4 x

106/ml in 1952 to 54.4 x 106/ml in 1972 (p<0.01). Although it is not clear in their

19

 

dc

cal

cltr
Alli

ml;

that :
numl
The r
78 mi
1111115

lump j

A stud
and)“

lime £11

description of the analysis, it seems likely the azoospermic men were included in this
calculation. Additionally there was significant deterioration in the fertility class (a
composite number reflecting semen volume, sperm count, perm mobility and sperm
morphology) from 1.95 to 3.70 (p<0.001). Men having semen analyses performed in their
clinic in 1972 did so as part of the routine work up before the fertility of the partner was
known, whereas in the 1950’s, the analysis was done only after a female problem was
ruled out. This change would actually favor an increase in sperm density over time in
their facility, whereas the opposite was found. This change may also suggest a change in
the technical sophistication of infertility assessments. A study in Scotland of 577
volunteer donors for a program in gamete biology compared sperm counts by birth
cohorts from 1951 to 1973, using a single laboratory with standardized methodology
(52). This study was notable for using donors unselected as to fertility status. They found
that a later year of birth was associated with a lower sperm concentration, a lower total
number of sperm in the ejaculate and a lower number of motile sperm in the ejaculate.
The median sperm concentration fell from 98 x 106/ml among donors born before 1959 to
78 mil/ml for those donors born after 1970 (p<0.002). The age of donors in the 4 cohorts
at first donation decreased with year of birth, but was not found to be independently

associated with sperm count in a multiple linear regression model.

A study from Paris (53)(n=1351) using fertile semen donors and one laboratory for the
analyses found a significant decrease in sperm concentration from 1973 to 1992. The
criteria for entry into the study was not specified and the age of the men increased over

time, although this did not completely explain the decrease. Another study performed in

20

 

 

 

Fr

)6

10

C0

lfp<t
1101c
Envj

d€n\

[I] «\U

1950

France, in the Toulouse area, investigated whether sperm densities had changed in 16
years, using a time series of sperm donor’s specimens from 1977 to 1992 at a sperm bank
of a university hospital (54). Donors had proven fertility, were within ages 20 to 45,
followed the recommended duration of abstinence of 3 to 5 days and thus were
comparable to the men participating in the study of Auger et al. They found no difference
in sperm counts. After discussing the similarities of the two studies, Bujan et al. suggest
that the observed differences in sperm count between the two studies at two different
locations may be due to environmental factors, including air quality, water supply and
life style. In support of this notion, a study in London was conducted comparing semen
parameters of partners of women who received gonadotropin therapy in 1978-1983 and in
1984-1989 (55). N 0 change was observed in mean age, clinical or socioeconomic status
during these two periods. A change in sperm density from 101 x 106/ml to 96 x 106/ml
was observed. However, when they later re-analyzed the data by dividing the couples
according to the source of their water supply, they found that the sperm density of men
living outside the Thames Water Supply Area changed from 99 x 106 to 110 x 106, while
the sperm density of men living inside the area changed from 105 x 106 to 76 x106
(p<0.05) (56). Significant changes in sperm motility (61.7% to 40.8%, p<0.05) were also
noted, but only for the men living within the area. These results suggest that
environmental factors, possibly present in the water, had adverse effects of sperm

density.

In summary, some but not all studies showed that sperm densities have fallen between the

19505 and the 1970s. Whether the decline has continued is uncertain, especially in the

21

 

Ma
thr
[ha
Car
and
CXpt
Inod
19131}
lht‘lr

adcc

USA. Carlsen et al’s review has a gap at this interface, and other models, which better fit
the later data, show an increase (41). Despite the valid criticisms leveled against Carlsen
et al’s study regarding population non-comparability and lack of methodological
consistency between studies, smaller studies that did not have these problems found
similar results. There seems to be valid geographical differences in sperm densities, and,
as suggested by some authors, this may be due to environmental conditions. Resolution
of this issue awaits future epidemiological studies using uniform methods aimed at

deciphering this relationship.

2)) Statistical methods
A major criticism of Carlsen’s et a]. data analysis was that they used inappropriate
statistical methods to analyze data from the studies that were not evenly distributed over
time ( 12 reports from 1930-1970, 48 from 1971 to 1990). Some critics have suggested
that lack of earlier publications prevents any sort of valid statistical conclusion (5) (41 ).
Carlsen et a]. Used linear regression to describe the relationship between sperm density
and year of publication. Subsequent re-analyses of their data have suggested that
exponential, logarithmic or cyclic models better describe their data (57) (41). All of these
models, however, show a decrease from the earliest times to the 1960’s. Carlsen et al. in
reply to these criticisms say they also tried the other models and found no better fit of
their data (45). They conclude that regardless of the statistical model used, the data show

a decline in sperm density over time.

22

 

\11

C0

Could

Another re-analysis of the Carlsen et al data, which included more recent publications,
found that the linear regression line was no longer useful in describing the data (58).
They also said that with the inclusion of more recent data, they did not find a decrease in
sperm densities since the 1960’s. On the other hand, yet another re-analysis of the
Carlsen et a]. data (59) using multiple linear regression models controlling for period of
abstinence, age, percent of subjects with proven fertility, specimen collection method,
study goal and location, found that mean sperm densities differed significantly across
countries with the highest densities found in Europe and lowest in non-Westem countries
(p<0.02). A decline in sperm density was seen in the USA from 1938 to 1988 (slope= -
1.50; 95% CI 1, -l.90 to —1.10) and in Europe from 1971 to 1990 (slope = -3.13; 95% CI,
—4.96 to —1.30), but not in non-Westem countries (slope = 1.56; 95% CI, -1.00 to 4.12).
Results from the non-linear models were similar to the linear models. They also found
significant regional differences often as large as the mean decline from 1938 to 1990.
The same authors published another analysis including additional data from 47 English
language studies published from 1934 to 1996 (60). With the additional studies their
results were consistent with those of Carlsen et a1. (slope = -0.94 vs -0.93) and their
previous results, suggesting that the reported trends are not dependent on the particular
studies included by Carlsen et a1. and that the observed trends previously reported for
1938-1990 are also seen in data from 1934-1996. The authors did not include information
on power analysis and did not provide the number of subjects in each study, so power

could not be determined.

23

 

CL

dt
0.:
hr:
011

Ca

dCL
ha\

are.

earl
real
POM

lime

Thus, the different statistical models all support a decline in sperm densities from the
earliest times to at least the 1960’ or 1970’s with regional differences playing an

important role.

The largest uncertainty in the data from the meta—analysis is whether the earliest sperm
densities, which predominately were from the USA and were high, are representative of
world-wide sperm densities. Since more recent reports from the USA still indicate fairly
high densities, especially those including data from New York (48) where the data in the
oldest reports in the meta-analysis were from, the degree of the decrease reported by
Carlsen et a1. is uncertain. Many of the studies included in the meta-analysis (47, 61, 62)
(33) and some of those published since then (52, 63) (64), have found significant
decreases in sperm density. A commonly held view is that decreases in sperm densities
have occurred over time, but are restricted to certain countries (46, 65), i.e. geographical
areas. Thus while sperm densities have decreased in areas of France (53), Denmark (61)
(64) and Scotland (52), they do not appear to have done so in Finland (66) (67). While an
earlier report from the USA suggests a decline (47), the most recent reports indicate
regional differences but no decline within those regions (49, 50). Several studies raise the
possibility that environmental factors play a role in the lowering sperm densities over

time (62) (54, 56).

2) Relationship between semen parameters and fertility
Semen quality has been used as a marker of male reproductive function in studies of

occupational health for several decades (9). However, the association of semen

24

 

 

L11"
c‘\
m

11’s

mOI
curt
dbnt
“lei
Fenij
110“}
nOlc(
”11me
prOPI‘Ir
31mm.

rephcd‘

parameters with fertility status has not been directly investigated. In fact there has never
been a population- based study on the possible association using a large, randomly
selected group of men. Two groups of investigators have used prospective study designs

to evaluate the association.

Bostofte and colleagues evaluated the association between measures of clinical fertility
and sperm density, motility and morphology in a prospective study of 1077 men
examined in The Sperm Analysis Laboratory in Copenhagen in 1950 and 1951(61). The
men were contacted 20 years later and asked to complete a questionnaire and 72.9%
responded. Fertility was defined as fathering a living child or as the number of living
children. They found that decreasing fertility correlated with decreasing sperm count,
increasing number of immobile sperm, poorer sperm motility and increasing number of
morphologically abnormal sperm (68). These parameters were all interrelated with
correlation coefficients from 0.36 to 0.67. Correlations between increasing number of
abnormal spermatozoa or decreasing sperm motility and decreasing chances of having a
live birth (p<0.01) and a longer time to pregnancy (p<0.01) were found (69) (51).
Fertility was impaired when 60% or more of the sperm were morphologically abnormal.
However men with poor motility still had a 32.7% chance of conceiving, so motility was
not considered to be a predictor of fertility. Correlations between sperm count and
number of living children and time to pregnancy were also observed (p<0.01).] The Cox
proportional hazard regression model was employed on 765 men whose semen had been
analyzed in 1950 and 1951, who had initially been diagnosed as infertile and who

replied to a questionnaire 20 years later. Age at semen analysis, percentage normal

25

 

pr
\x‘l

“Ll

C on
labo
incrc
for p
0.56 I
105/tr

the lil

Sperm
inCTt’d}
mOI‘Pht
pmpOn

“Ere lllt

sperm, percentage mobile sperm and the degree of motility were all found to be

significant independent predictors of individual pregnancy probability (70).

More recently Bonde et a1.(7 1) studied the association between semen quality and the
probability of conception in a single menstrual cycle (fecundability) in Danish couples
with no previous reproductive experience. Couples (430), aged 20-35 years, one of which
was a members of a Danish trades-union, were followed for 6 months or until a
pregnancy occurred. Semen samples were collected at the start of the study and during
the menstrual period of each menstrual cycle. Sampling after 3 days of abstinence was
encouraged but not demanded and all analyses were carried out by two laboratories.
Comparison of the variation in sperm counts of the same sample (n=28) by the two
laboratories was not significant (p=0.82). They found that the probability of conception
increased with increasing sperm concentration up to 40 x 106/ml. The odds ratios (ORs)
for pregnancy with increasing sperm concentrations ranged from 0.30 (95% CI, 0.16-
0.56) for counts of 0-9 x 106/ml to 0.96 (95% CI, 0.68-1.38) for counts of 40—79 x
106/ml. A threshold of about 40 x 106/m1 above which there was no significant increase in

the likelihood of pregnancy was obtained using a logistic regression model.

Sperm morphology was also a significant determinant of pregnancy. There was a linear
increase in the likelihood of pregnancy with increasing proportions of sperm with normal
morphology from 10% to 60%. A positive correlation with sperm concentration and
proportion of normal sperm was found (r=0.40, p<0.0001), although both parameters

were independently associated with likelihood of pregnancy. The likelihood of pregnancy

26

 

tn

pn
11m
adh
C0111
their
dillx

10:31,

In a 5

Ubt‘d (
1111011 S
were (1
0116 39

11,1 11am

significantly decreased if the proportion of non-motile sperm was greater than 70%,
although sperm motility was not a significant factor in a logistic regression model
containing sperm count and morphology. In the analysis of menstrual cycle-specific
semen samples, the likelihood of pregnancy increased from 8% in cycles with sperm
counts below 10 x 106 to 25% in cycles with counts greater than 40 x 106. The adjusted
OR for pregnancy in a menstrual cycle with sperm concentration increasing at 1x106/m1
increments was 2.45 (95% CI, 1.53-3.89, p=0.0001). This study provides a much needed
assessment of the association between semen quality parameters and a measure of
fertility using a sample that is unbiased with regard to fertility potential, which has been a
problem in many studies on sperm count. The selection of trade union members,
however, may make it less representative of the general population. The lack of strict
adherence to a 3 day period of sexual abstinence may make their measurement of sperm
concentration less valid, and the authors mention that the mean sperm concentration in
their study increased by 5.2 x 106 sperm/ml per day of sexual abstinence up to seven
days. However, they mention that addition of duration of sexual abstinence in their

logistic regression model did not change the likelihood of pregnancy.

In a study to evaluate infertility according to shifts in sperm counts, Bonde et a1. (64)
used data from 10 separate Danish occupational semen studies conducted from 1986
through 1995 to establish a baseline distribution of sperm counts. Although all the studies
were designed to evaluate occupational exposures, all but two studies found no effect.
One semen sample, collected by masturbation, was evaluated within 2 hours of collection

by trained technicians. The median sperm count was 56 x 106/ml with a range of 0 to 504

27

 

HM

tat
int:

oft

1n 5!
10 p1
Add
cont;
COny

Pdfar

FUHh‘
a1111011

x 106/ml. Using this data and data from the study described above (Bonde et al), the
authors constructed additive or multiplicative models to create alternative distributions
with median sperm count values changed by 25-100%. The resulting models suggest that
the relationship between sperm count shift and fecundability depends on the median
levels of the sperm count and the type of shift. Variables interacting in a multiplicative
model tend to have minor effects on fecundability while those with additive interactions
can decrease fecundability by 30% and fertility by 70%. Thus, depending on the
variables and their type of interaction, for example exposure to dioxin and DDT
interacting additively, it is possible that changes in sperm counts may be an early warning

of changes in fertility.

In summary, these studies have found an association between measures of fertility (time
to pregnancy, live birth) and one or more semen parameters, including sperm counts.
Additionally, a theoretical framework for the interactive effects of environmental
contaminants on sperm counts has been developed, and stresses the importance of
considering multiple exposures in designing studies to evaluated changes in semen

parameters.

b. Changes in the incidence of abnormalities of the male genital tract
Further support for a decline in male reproductive health comes from studies on
abnormalities of the male genital tract, including cryptorchidism and hyposapdias, and on

testicular cancer.

28

 

Av

hull

JDOU
due 1

tendt

Assessment of trends in the incidence of cryptorchidism and hyposapdias has been
hampered by the lack of consistent measuring techniques and strict definitions of both
conditions. An additional problem for diagnosing cryptorchidism is the age-dependent
retractability of the testes leading to misclassification of both undescended and descended
testes. It has also been reported that young boys with retractile testes that would have
descended with time may have been diagnosed and operated on for undescended testes at
an early age to prevent sterility (72). To compensate for this problem, a strict definition
of cryptorchidism that eliminates the possibility of retraction should be Used, i.e. a testis
that cannot be drawn at least 4 cc below the pubic crest at some time before 1 year of age
(73). Additionally the evaluations for cryptochidism should be performed at birth, at 3
months and at one year. Problems in diagnosing hypospadias include the existence of
forms considered minor (distal, glandular or coronal) and severe. Minor forms occur
about 75% more frequently (74, 75). The rising trend for both these anomalies may be
due to more frequent or earlier diagnosis of the minor forms over time or to an increasing

tendency to report them to congenital anomaly registries (76).

a. Cryptorchidism
Three large prospective studies on the incidence of cryptorchidism, all adhering to the
same standards, have been published and can be compared. The first study on 3500
infants born in a hospital in London in the 1950’s used very accurate measurements of
testes positions and a strict definition of a cryptorchid testis and has thus served as a
reference for subsequent studies (73). This study found that at birth, 2.7% of full term

male babies weighing 2500g or more had one or two undescended testes, while 21.0% of

29

tint

sin

TRIO
same
HOspj
bola
zfiutho
1ESsrh
10250

re‘PECI

infants weighing less that 2,500g had the same condition. At three months the incidence
of cryptorchidism in boys with birth weights equal to or greater than 2500 g or less than
2500 g were 0.91% and 1.74%, respectively. Another study from England conducted in
the 1980’s on 7441 infants found the incidence to be 5.2% for infants less than 2500 g
and 1.61% for those over 2500 g at 3 months (77), indicating an increase in incidence

since Scorer’s study.

A study from England in 1984 examined the Hospital Inpatient Enquiry data for England
and Wales from 1962 to 1981 and found that the diagnosis of undescended testes had
increased 2.3-fold (72). The cumulative discharge rate up to age 15 for undescended
testes increased from 1.4% in 1952 to 2.9% for 1977, adjusted for double admissions.
There are several problems with this data including a variable definition of
cryptochidism by different physicians and the inclusion of subjects who were cryptorchid

at birth but whose testes descended by one year of age.

Two more recent studies, one from England and one from New York, both using the
same technique and definition as Scorer found contrasting results. The John Radcliffe
Hospital Cryptorchidism Study Group (78) examined cryptorchidism at birth in 7441
boys, born from 1984 to 1988, whose mothers lived within the Oxfordshire Health
Authority in England, and if present, at 3 months of age. At birth 22.8% of boys weighing
less that 2500 g were cryptorchid while only 4.1% of boys weighing greater than or equal
to 2500 g had the same condition. At 3 months, the values were 5.2% and 1.61 %,

respectively. These latter values represent a 197.4% and a 77.4% increase compared to

30

 

[0
int

v.11

51“?

the values found by Scorer at 3 months. The study from New York examined 6935
consecutive male babies delivered at Mount Sinai Hospital in New York City from 1987
to 1990 (79). At birth 19.8% of babies less than 2500 g were cryptorchid while 2.2% of
infants weighing greater than or equal to 2500 g were cryptorchid. At three months these
values dropped to 1.94% and 0.91%, respectively. These values are close to the values
reported by Scorer and suggest no change had occurred since Scorer’s study. The New
York cohort, however, was very diverse ethnically and racially and may have differed
significantly from the British groups. Overall these data suggest an increasing trend in
cryptochidism rates, although it is difficult to draw conclusions from these different

populations.

2) Hypospadias
Accumulating evidence points to an international increase in the prevalence of
hypospadias. Based on data from national registers, a steady increase in the incidence of
hypospadias has been observed in England and Wales (80), Hungary (Czeizel 1985),
Sweden (Kalen 1982; 1986), Norway (Bjerkedal et a1 1975) and Denmark (Kallen 1986;
WHO). The largest increases occurred from the 1970’s to the late 1980’s (WHO
Clearinghouse for malformations, 1991). In England and Wales, the incidence of
hypospadias increased from 7.3 per 10,000 births in 1964 to 16 per 10,000 births in the
early 1980’s and then decreased slightly to 11.7 per 10,000 in 1990. Data from Hungary
followed a similar trend, rapidly increasing from 5.5 per 10,000 in 1964 to approximately
23.9 per 10,000 in the early 1980’s where it has remained. Data from Denmark indicate

an increase in incidence from 1970 to 1981 from approximately 7.5 to 12 per 10,000

31

 

In Cl
ALN
11.1110
1986
preta
lhlt‘tu
pm)“
Ifglst
bend

butni

I“the
Auunr
Binh l
ILrCI

a““Ual

(Kallen 1986). An additional increase occurred from 1982 to 1988, but may have been
influenced by a change in the registration systems. Sweden also had an increase in the
early 1970’s: a 40% increase was noted from data collected in 1965 to 1968 compared to
that collected in 1974 to 1982 (Kallen 1986). Finally in Norway, the prevalence of
hypospadias at birth increased from 7-8 per 10,000 between 1967 and 1971 to 13 per

10,000 in 1973 (Bjerkedal) and to 20.7 per 10,000 in 1988 (WHO 1991).

In contrast, an increasing trend was not observed in Finland, Spain, New Zealand,
Australia or Czechoslovakia (WHO 1991). A more recent study from Finland used their
national hospital discharge registry birth defects data for boys born between 1970 and
1986 to find the number of boys treated for hypospadias by the age of 8 years (81). A
prevalence of 28.1 cases per 10,000 live male births was found and it remained constant
throughout the study period. The rate, however, was approximately 3-fold higher than
previously reported, perhaps because of the variation in completeness of birth defects
registry data used in the earlier studies. Thus in Europe and Scandinavia an increasing
trend in the incidence of hypospadias was observed from the 1960’s through the 1980’s,

but may have leveled off (82) (83).

In the USA the prevalence of hypospadias, determined from data from the Metropolitan
Atlanta Congenital Defects Program (MACDP) for 1968 to 1993 and the nationwide
Birth Defects Monitoring Program (BDMP) from 1970 to 1993, increased (75). In the

MACDP database, the total hypospadias rate nearly doubled between 1968 and 1993. An

annual rate of increase of 2.9% obtained by using linear regression and was significant

32

 

'JI

trnr
fnwn
rate

111“

TR 1)
sum
CODE
S}\‘Ic
inure.
dlxcn
hwos

f0”hs

1116 ct
RWCQ
"K“Cat
WUHdV
Ix 111cm

Count”

(p< 10°). The overall increase occurred at a rate of 1.4% per year among Caucasians and
5.7% among non—Caucasians. The rate of severe hypospadias increased three-to five-fold
from 1.1 per 10,000 live births in 1968 to between 2.7 and 5.5 per 10,000 births per year
from 1990 to 1993 (p for trend <10'6). Data from the BDMP database showed that the
rate for hypospadias nearly doubled from 1970 to 1993, going from 20.2 to 39.7 per

10,000 (p for trend < 106). All four regions of the US showed an increasing trend.

Two more recent studies arrived at conflicting conclusions. A study using two US
surveillance systems also found an increase in prevalence (84). A study USing the
Congenital Malformations Registry and State Wide Planning Research Cooperative
System of New York State for the incidence and repair rates of hypospadias found no
increase in incidence from 1983 to 1995 in New York State (r=-0.225, p=0.4)(85). The
discrepancy between these two reports may in part be due to misclassification of
hypospadias by Canning (84) since they did not differentiate between minor and severe

forms.

3) Testicular cancer
The evidence supporting an increase in the incidence of testicular cancer is stronger than
for cryptorchism and hypospadias at least in part because it is easier to diagnose. Studies
indicate that the incidence of testicular cancer has increased for the last several decades
worldwide (reviewed in (86)). Reports from cancer registries indicate that the incidence
is increasing in England and Wales (87) (88), Scotland (89), the Nordic and Baltic

countries (90) (91), Australia (92), New Zealand (93) (94), northern Europe (91 ), Canada

33

 

A rec
I951.

dguin
In On
Title It
1001:).
100.11
With a

The h

(95) and the USA (96, 97). A USA study using the Connecticut Tumor Registry for 1935
through 1979 found Cln epidemic increase over time in the risk of testicular cancer for
men aged 15-44, with the most recent birth cohorts showing the greatest rate of increase
(96). A subsequent study using the same database for 1935 to 1992, found that the
incidence rate of testicular cancer (all histologies) has increased 3.5-fold for Caucasians
during the past 60 years: 1.46 per 100,00 in 1935-1939 to 5.01 per 100,000 in 1990-1992
(97). The rates for seminomas and non-seminomas increased since the 1950’s: the age-
adjusted incidence rate for seminomas increased from 1.03 per 100,000 men in 1950-
1954 to 2.60 per 100,000 in 1990-1992, while that for the nonseminomas increased from
1.14 to 2.41 for the same time period. For seminomas the largest increase occurred for
men 20 to 44 years old, while the rate for nonseminoma was highest for men aged 15 to
34. The increases for men born after 1910 for both types of cancer were mainly

explained by a strong birth-cohort effect.

A recent study from Canada also found an increase in testicular cancer incidence rates
(95). The study from Canada involved 7296 incident cases of testicular cancer, divided
again into seminomas and non-seminomas, which were recorded in the cancer registries
in Ontario, Saskatchewan and British Columbia from 1970 and 1995. The age-adjusted
rate for seminomas increased by 53% from 1.5 per 100,000 males in 1970-1971 to 2.3 per
100,000 in 1994-1995. The non-seminomas increased by 91% from 1.1 to 2.3 per
100,000 during the same time. Non-seminomas occurred in younger men, aged 15-19,
with a four-fold increase. Both types of testicular cancer showed a birth cohort pattern.

The high occurrence of this cancer in young men compared to men over 50 has been

34

interpreted to mean that exposure to risk factors early in life, possibly in utero are more

likely to be important than exposures in adulthood (91).

4) Possible associations between sperm counts, cryptorchidism,
hypospadias and testicular cancer

There are marked racial and geographic differences in the incidence of testicular cancer,
some of which parallel the reported decreases in sperm count. For example Denmark has
one of the highest incidences of testicular cancer (7.8 per 100,000) (90) (91) and has
reported a decrease in sperm counts (28). On the other hand, Finland has a low incidence
of testicular cancer (1.3 per 100,000) (91) (90) and has reported little change in sperm
counts (98) (67). The etiology of testicular cancer is largely unknown, but cryptorchidism
is the only established risk factor and the risk is reported to be increased for the
descendend testis of monorchids (99). Additionally several prenatal risk factors are
shared by both, including high levels of estrogen in the first trimester (100), premature

birth (101), and hypospadia (78), suggesting that they may share etiological factors.

Based on these observations some investigators argue that changes in sperm counts
should not be viewed as an isolated phenomenon, but as part of a larger decline in male
reproductive health (102) (103) (86) (104) (105). They argue that the possible decline in
sperm counts, the increased incidence in testicular cancer, as well as the increasing
incidence of cryptorchidism and hypospadias result from adverse prenatal and postnatal
effects on the testis. While it is obvious that the congenital anomalies of cryptorchidism

and hypospadia have prenatal origins, it has been suggested that testicular cancer arises

35

 

III

11']

10

m

0 l"

\C\

from cells of carcinoma in situ, which are presumed to be malignant primordial
gonocytes (106). The observations that lower sperm counts are associated with damage
to the testis (107), testicular cancer (108) and a cohort effect (52), suggest a possible link
with a perinatal exposure. It has been hypothesized that in utero exposure to high levels
of estrogen or other hormonally active chemicals (86) or to low levels of testosterone

(104) is responsible for the damage.

Studies on the effects of diethylstilbestrol (DES), a nonsteroidal estrogen administered to
several million women between 1950 and 1970 to prevent miscarriage, both support and
refute this hypothesis. In case-control studies, male offspring of DES-exposed mothers
had an increased risk for hypospadias and cryptochidism [Gill , 1979 #837], although
their fertility was not impaired ( 109). Several other case-control studies have found an
association between DES exposure and testicular cancer, while others have not (reviewed
in (l 10). Recently, a study reviewed and combined data from 4 prospective cohort studies
on in utero DES exposure and testicular cancer (Mayo Clinic cohort, Dieckmann cohort,
Women’s Health Study cohort and Home cohort) (1 10). The study population included
3613 men followed for 16 years. They concluded that testicular cancer rates among DES-
exposed men may be elevated since the found a R of 3.05 (95% CI, 0.65 to 22.0), which
was not statistically significant. The increased testicular cancer risk was limited to men
in the Mayo Clinic cohort, which suggests that this cohort may differ in some way from
the others. The authors note that the men within the Mayo Clinic cohort were exposed to
lower DES levels than men in the other three cohorts. Animal studies have found a

negative effect Of a low but not a high DES dose on prostate abnormalities (11 1).

36

 

 

 

 

 

 

 

 

 

 

ct

_ 1
CV

An
Pin;

Csli

RCC:
heal
pro;

0.51

Additionally the current study had only a 30% power to detect a statistically significant
effect due to the low frequency of testicular cancer in the study population. Thus, this
report lends marginal support to the hypothesis that increased levels of estrogenic
hormones during fetal development have adverse effects on male reproductive
development. Thus the role of endocrine disruption on male reproductive problems

remains controversial (1 12) and awaits further investigation.

Another critical period in development is during adolescence when sex hormones again
play an important role. It has been suggested that exposure to estrogen or chemicals with
estrogenic or antiestrogenic properties during this time may adversely affect
development and function of the testes. A variety of substances found in the environment
have these properties as well as others that make them candidates for reproductive

toxicants.

c. Changes in the sex ratio
Recently the stability of the sex ratio has been proposed as a marker of the reproductive
health of a population (1 13). The sex ratio of a population, usually expressed as the male
proportion (male births/total births), is usually stable over time and in most countries is
0.512 (1 14). A change in the sex ratio has also been viewed as a sensitive marker of
exposure to toxic chemicals (1 15) and a sentinel event for changes in cancer risk (1 16).
Several surveys have shown a decrease in the sex ratio in the last 20 to 40 years in the
USA (1 17), Canada (1 18), Latin America (1 19), Denmark (120), Netherlands (121),

Germany (122), England and Wales (123) and Japan (124). An increase in the sex ratio

37

.-..J

 

“2

(311

an.
191
in!
bir‘.
inc:
lllf

1W1

Orin:
Comte
COHIH

dlen‘

was noted in Italy from 1930 to 1989 and was speculated to be due to improvement in

environmental conditions (125).

In the USA the sex ratio of live births, taken from birth certificates of live born infants
and compiled by the US National Center for Health Statistics, declined from 0.513 in
1969 to 0.512 in 1995 for all live births (1 17). A change in the sex ratio of 0.001 results
in the loss of one male birth per 1000 live births. The decline however was seen for white
births (OR, 0.9935; 95% CI, 09919-09952) while the ratio for black births actually
increased (OR, 1.0208; 95% CI, 1.0162-1.0254). Smaller regional changes were observed
for white births. A greater decrease was reported in Canada where the sex ratio declined
from 0.5147 in 1970 to 0.5125 in 1990 (p<0.001) (1 l8), representing a loss of 2.2 males
per 1000 live births. In these two countries there is no reason to suspect a reporting bias
in favor of female births. Thus it appears that almost all reports indicate a decline in the

sex ratio.

2. Exposure to OCs and mercury
a. OCs and mercury in the environment: natural history and trends
1) OCs
a) Natural history
Organochlorine compounds found in the Great Lakes have long been a public health
concern. Those found at high enough concentrations historically and currently to be of
continuing interest include polychlorinated biphenyls (PCBs), polychlorinated

dibenzodioxins (PCDDs), specifically 2,3,7,8-; tetrachlorodibenzo-p-dioxin (TCDD;

38

 

diox.
\pCCl
IDdh‘
13I)E
chbc'

chcni

‘g

dioxin), polychlorinated dibenzofurans (PCDFs) and dichlorodiphenylchlorethanes,
specifically 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane (p,p’-DDT; DDT) and its
major and persistent metabolite l,1—dichloro-2,2-bis(p-chlorophenyl)ethylene (p, p ’-
DDE). They have been found in Great Lake fish and have been shown to have adverse
effects on reproduction in a variety of species. The diagram below illustrates their

chemical structure.

Dichlorodiphenylchloroethanes

@fii-QD} @71‘©'C‘

01- --C—01
. Cl Cl
p,p'-DDT p,p'—DDE
Polychlorinated Biphenyls
3 2 2' 3‘
' Cl Cl
4 4'
O G
5 6 6‘ 5‘
P085 .. general structure 3,3',4,4‘-totrachlorobiphenyl
. a coplanar PCB
Polychlorinated Dibenzofurans
2 1 9 8 Cl Cl
4 6
O o
PCDFs -- general structure 2,3,7,8—tetrachlorodibenzofuran

Polychlorinated Dibenzodioxins
1 9
2 O 8 Cifiomc'
3 o 7 Cl 0 Cl
4 6

PCDDs - general structure 2.3.7.8-tetrachlorodibenzodioxin
dioxin

Figure 1. Structure of PCBs, dioxin, dibenzofurans

39

 

(X:

Grou:
'Ihelr
l‘lng\‘
I 123 I,
adhC\
0rSkim
lldnq}

mafiuf

OC can be broadly classified as pesticide and non-pesticide (126). The non-pesticide
OC’s, PCBs, dioxin and PCDFs, belong to the family of polyhalogenated aromatics and
are man-made (127). PCBs were manufactured by the addition of chlorine atoms to the
two biphenyl rings. The process yields variable levels of chlorination, producing a
theoretical 209 congeners depending on the number and location of the chlorine atoms on

the biphenyl rings (Figure 2).

meta ortho

 

 

meta ortho

 

Figure 2. Positions of chlorines on PCBs

Groups of congeners with the same number of chlorines are referred to as homologs.
Their stability depends on the number and position of the chlorine atoms on the biphenyl
rings, with the higher chlorinated congeners possessing greater stability and persistence
(128). Because of their chemical stability, PCBs were used as hydraulic fluids,
adhesives, plasticizers in paint, heat transfer fluids, wax extenders, dedusting agents,
organic dilutents, lubricants, flame retardants and as dielectric fluids in capacitors and
transformers (127). It is estimated that approximately 1.4 x 109 pounds of PCBs were

manufactured in the USA from 1930 to 1975. The commercially manufactured PCB

40

 

111

Ill

ur
ha

ha

ch.

£1111

PC

dc;

mixtures were marketed under trade names, such as Phenoclor, Kanechlor, Clophen, and,
in the USA, Arochlor (127). Different Arochlor formulations were produced, containing
mixtures of congeners and thus variable levels of chlorine ranging from 21% to 68%.
Although they were first produced in the 1920’s, significant production did not occur
until the 1950’s. Due to their detection in the environment production of PCBs was
banned in the USA in the late 1970’s. As a consequence of atmospheric transport, PCBs
have been detected in rivers, lakes and ocean sediments and their biota. Through the
bioaccumulation of PCBs in water and soil sediments, they bioconcentrate in the food
chain, with the highest levels found in fatty tissues of carnivores such as the herring gulls

and mink.

PCDFs, which can also exist in various forms (approximately 135 congeners) again
depending on the chlorination degree and pattern, are by-products of the PCB
manufacturing process. Thus the commercial Arochlors actually contained a complex
mixture of isomers and congeners of PCBs and PCDFs and were the major source for the
introduction of these compounds into the environment. PCDDs (dioxins) are also by-
products of industrial syntheses, such as in the production of 2,4,5-trichlorophenol and
bleaching processes, and exits as 75 congeners again depending on the position and
degree of chlorination. The most toxic and most widely studied PCDD is 2,3,7,8-
tetrachlorodibenzo-p-dioxin, which is commonly referred to as “dioxin”. It is the most
toxic synthetic chemical known. It is important to emphasize that all human and
environmental exposures to these compounds involve complex mixtures that interact

additively, synergistically or antagonistically (129).

41

 

Cht
\66
mi
salt
C017
60!
C1111

littl

The
m aj .

1Il\'t]

See 5
EICIIV
rm:
C9mm
Congt

hil10gi

Dioxin and certain PCB congeners, although not structurally related, share a common
chemical conformation. PCB congeners that lack chlorines in the ortho position (please
see Figure 2), but have them in both para and at least 2 meta positions on the biphenyl
ring, are referred to as coplanar since the two rings are hypothesized to rotate into the
same plane giving the molecule a flat configuration. These congeners are generally
considered to be the most toxicologically active based on their ability to induce CYP
enzyme activity via the Ahr (129). Those congeners with a single ortho-substituted
chlorine are considered to be less toxic while the lower chlorinated congeners are possess

little or not activity (129).

The major, although not sole (130) (131), pathway thought to be responsible for the
majority of dioxin’s toxicity involves induction of cytochrome P450 (CYP) genes
involved in activation of environmental toxicants. Both dioxin and the coplanar PCB
congeners initiate this pathway by binding to the aryl hydrocarbon receptor (AhR) (132).
Once these compounds bind to the AhR, the complex translocates to the nucleus and acts
as a transcription factor activating a variety of genes, including P450 genes ( 133) (please
see section B.4. a. l) for more details). The toxicity of a compound that has dioxin-like
activity, such as the coplanar PCBs, is often expressed as its toxic equivalency factor
(TEF) with dioxin as the reference value (TEF=1). The TEF model for PCBs assumes a
common mechanism of toxicity and additivity for the toxic effects of the individual

con geners. The TEF value assigned to a specific PCB congener, as well as other

halogenated aromatic compounds, is based on results of assays measuring binding to the

42

AhR, long term carcinogenicity, immunotoxicity, acute lethality, and enzyme induction,
and often requires a subjective assessment of the results (134); (126). Additionally, both
dioxin and the coplanar PCB congeners have weak anti-estrogenic activity (135) ( 136)

(137).

The moderately chlorinated congeners, with 6 or 7 chlorines, have high TEF’s, are the
persistent in nature and are found most often in human tissues. If they are not chlorinated
in the ortho positions, they are usually unable to bind to the AhR and lack the associated
activities. These congeners have been reported to have neurotoxic, carcinogenic and
endocrine disrupting activities (138). The lesser chlorinated congeners are readily taken
up by organisms but also readily eliminated. Those more heavily chlorinated congeners
(those having seven to ten chlorines) occur in low concentrations in the environment,

remain tightly bound to soil and sediment and are thus less bioavailable ( 129).

The lower chlorinated coplanar and ortho-substitued PCB congeners also have biological
activities. The lower chlorinated forms have lower TEF values and have weak estrogenic
activity in various in vitro assays (135). The ortho-substituted non-coplanar PCBs

con geners, which have low affinity for the AhR, have biological activities with potential
toxicity (139). Some of these congeners, such as 2,2’,6,6’ tetrachlorbiphenyl with two
chlorines in the ortho position, have been shown to have neurochemical effects and are
weakly estrogenic. It is this group of congeners (140) (141) that appear to be partly
responsible for the behavioral changes and learning deficits that have been observed in

children exposed in utero to PCBs through fish consumption (142) ( 143). Investigations

43

 

 

01 1
(db

dc‘ll

Thu~
c\pu
nctc
rcprc
mcnt
Chku
PCB
have
PCI3
91cc
“Dbh
that d
Shnh.
throué

Wirth.

of the effects of Archlor 1242 on release of reactive oxygen species from neutrophils
(discussed in section 4. b.), has revealed that ortho, but not coplanar PCB congeners
active several intracellular pathways required for super oxide anion (02') production
(144). These PCB congeners are more prevalent in the environment than the dioxin-like
congeners, possibly due to biotransforrnation of higher chlorinated and coplanar species

(139).

Thus, it is becoming clear that analyzing associations between a health outcome and
exposure to total PCBs may mask important effects mediated by specific congeners. The
net effect of exposure to PCBs, through fish eating, for example, is thus likely to
represent additive effects as well as opposing effects of the different congeners. As
mentioned above, higher chlorinated PCBs have anti-estrogenic effects while the lower
chlorinated congeners have estrogenic effects. It is thus important to use a technique for
PCB analysis that will detect all possible congeners present in a specimen. Studies which
have reported PCB exposures in terms of the specific congeners have found that different
PCB congener profiles are associated with different exposures. For example, the profile
of congeners observed in fisheaters (145) is quite different than the profile observed in
ambiently exposed individuals (146). Occupational exposures yield a congener profile
that differs from that of individuals exposed environmentally through diet (147).
Similarly, the congener profile of fisheaters differs from that of firemen exposed to PCBs
through smoke inhalation or from a mechanic working with hydraulic fluids (Mullard and

Wirth, unpublished data).

44

DDT, a pesticide OC, was synthetically produced and widely used in agriculture and
forestry from the 1940’s to the 1960’s. DDT preparations contain a mixture of two
isomers, p, p ’-DDT and o,p ’-DDT. P,p ’-DDE, the major break down product of p, p ’-

DDT, is the most frequently found form in human tissue due to its long half life.

Since it is likely that different OCs as well as different congeners of these OCs are found
in the same locations and will bioaccumulate, most environmental exposures will involve
a mixture of these compounds. Several studies have shown that OC can interact in vivo.
Co-administration of dioxin and a non-dioxin-like PCB (congener 153) to rats resulted in
a strong synergistic effect on porphyrin accumulation in the liver, an effect mediated by
AhR-induced activation of CYP genes (CYP1A2) ( 148). Administration of dioxin alone
or several concentrations of PCB 153 alone had no effect (1.4-3.0 ug/g liver). Dioxin
plus PCB 153 at the lowest dose increased porphyrin levels to 300 ug/ g liver and to 1223
ug/g liver at the highest level. DDT metabolites found in the heavily contaminated Lake
Apopka in Florida were found to interact synergistically with isolated alligator oviduct
estrogen receptors (aER) at concentrations measured in contaminated alligator eggs
(149). P,p’-DDE and p,p’DDD, DDT metabolites, both bound to the alligator ER, and in
combination decreased estradiol binding in a greater than additive manner (p,p’-DDD,
94% decrease; p,p’-DDE, 93% decrease; p,p’-DDD plus p,p’-DDE, 74% decrease;
p<0.05 compared to no chemical control and individual metabolites). Similar results have

been reported for other pesticides interacting to decrease estradiol binding to the human

45

ER and to increase human ER—mediated transcription of an estrogen-sensitive reporter

gene in a synergistic manner (150, 151).

Administration of several PCB congeners to rats and mice increased hepatic TCDD
receptor levels (Safe, 1986). Pretreatment of these animals with the PCB congeners
followed by TCDD markedly increased hepatic AHH and EROD induction, activities

regulated by the AhR.

OC can also interact antagonistically. Several halogenated aromatic compounds have
been shown to antagonize the effects of TCDD (134). Aroclor 1254, a weak AhR agonist
can partially antagonize several TCDD-mediated responses, including induction of AHH
and EROD activities and immunological responses in vitro. A dibenzofuran congener,
(1,3,6,8—TCDF) competitively antagonized binding of TCDD to the AhR in vitro

(reviewed in (134)).

The observation that OC interact in both synergistically and antagonistically complicates
risk assessment as well as attribution of causality to any single chemical or congener.
Analysis of specific contaminants in an environmental exposure at least identifies the
individual chemicals of concern, and combined with current knowledge of their effects,

can offer a crude explanation for the observed effects.

b) Trends

46

Since PCB production ceased in 1977, reports indicate that the levels of PCBs found in
the Great Lakes basin have generally declined (152) and will continue to decline although
at a slower rate (153). According to the US. EPA, however, the concentration of PCBs in
this “new equilibrium” is well above safe water quality standards (154). As recently as
April, 2000, the EPA reported in a press release that PCBs still leak from Lake
Michigan, and probably into other sites in the Great Lakes (155) providing new sources
of potential human exposure. Such contaminant input from point sources, spills and
direct run off from urban or rural areas will affect local concentrations and hence local
exposures but will contribute little to the overall contaminant burden, which for PCBs in

Lake Michigan is estimated to be 80,000 kg ( 156).

2) Mercury : natural history and trends in the environment
a) Natural history
Mercury is a toxic heavy metal that exists in trace amounts in the earth’s crust. Various
natural processes, including volcano eruptions, weathering of rocks and forest fires
release inorganic mercury into the atmosphere where it is dispersed worldwide. Mercury,

in dry or wet form, is deposited back to lakes, rivers or oceans.

Anthropogenic sources of mercury contribute significantly to the levels of mercury found
in the environment. Elemental mercury is used in thermometers and dental amalgams and
inorganic mercury is used in manufacturing processes and is usually present in hazardous
waste sites. However, while exposure to these forms of mercury is associated with human

health hazards, such as renal toxicity, these types of exposures tend to occur after

47

accidents. Otherwise exposure to these forms of mercury occur at low concentrations and
is not a major public health concern ( 157). However, through a process of methylation,
mediated by bacteria species found in soil, sediments, fresh water and salt water, mercury
forms stable complexes with organic compounds. The newly formed methylmercury,
which is the most toxic form of mercury (158), along with methyl mercury from
pesticides and mildew control agents, enter the aquatic food chain (159). In water bodies,
it can be taken up by fish and bioaccumulate in the aquatic food chain biomagnifying to

tens of thousands to millions of times the concentration found in water (160) (157).

b) Trends
Since pre-industrial times, background concentrations of mercury in the atmosphere have
increased two to five times (157) due to new human sources of mercury release. These
new sources include burning of mercury-containing fossil fuels, release of mercury-
containing by-products from manufacturing processes (chloralkali plants, incinerators)
and the release of mercury from paints, broken fluorescent lamps, switches,
thermometers, and mercury-containing batteries (158) (161). In 1996, the background
concentration of mercury in the atmosphere over the oceans was estimated to be 1.6
nanograms per cubic meter (162) whereas in pre-industrial times, it was estimated to
have been 0.5 ng/m3. Additionally the deposition of mercury onto the surface of the earth
has also increased from approximately 3.7 11ng in 1850 to 12.5 rig/m2 currently (160).
Approximately 25% of the atmospheric mercury deposited over mid-continental North
America has ended up in lakes. A modeling analysis conducted by the USEPA (160)

predicted that the southern Great Lakes region will experience further high rates of

48

mercury deposition. Even small increases in atmospheric mercury has the potential to

yield measurable increases in methymercury levels found in aquatic organisms (161).

b. OCs and mercury levels in fish
1) OCs
Due to their lipophilic nature, OC contaminants tend to accumulate in the fatty tissues of
fish. Thus “fatty” fish, such as salmon and trout, tend to have higher contaminant levels
than leaner fish such as perch and bluegills. Additionally the size of the fish, which in

part is a function of its age, affects its contaminant burden.

The Michigan Department of Environmental Quality-Surface Water Quality Division
(MDEQ-SWQD) monitors the fish levels of environmental contaminants commonly
found in waters of the Great Lakes and inland water bodies. The monitored contaminants
include PCBs, DDT and DDE, dioxin, other organochlorine pesticides and mercury (see
Appendix B for a summary of chemicals quantified in fish). MDEQ-SWQD samples fish
from 40 sites statewide and issues a report from which the Michigan Department of
Community Health (MDCH) bases their fish consumption advisories for specific fish in
specific locations. The number and species of fish monitored per site varies and the
specific sites evaluated from year to year varies. The number of fish per species varies
from 2 tolO, with the majority over 5 fish. It is thus fair to say that the contaminant
information for fish species and site per year are inconsistent, making it hard to
generalize. When the permissible level of a particular contaminant is exceeded for a fish

species (the “trigger level”), MDCH issues a fish consumption advisory. The “restrict

49

consumption” advisory is issued when more than 10% of a particular species of a given
length exceeds the trigger levels, indicating that that there is a limit on the number of
meals of that particular fish that can be eaten without possible health risks. Beside
advisories for specific water bodies, MDCH also issues statewide advisories covering
certain predator fish, such as northern pike, walleye, small mouth bass and large mouth

bass, from inland lakes and reservoirs.

In their 1998 Michigan Fish Contaminant Monitoring Program Annual Report (163),
MDEQ-SWQD found that contaminant concentrations were above the MDCH’s trigger
levels in all fish collected from 35 of 40 sites. (See Appendix B for a list of the trigger
levels). For total PCBs, fish from 78% of the sites exceeded the women and children
trigger level, while 11% of the sites exceeded the general population trigger level. At
least one species from 19 of these 21 sites was under a fish a consumption advisory. For
TCDD fish from 50% of the sites exceeded the MDCH trigger level and species from all
of the sites were under a consumption advisory. Few sites reported fish with levels of
DDT levels above the trigger level. An exception was all fish species from the Pine
River, St. Louis impoundment, which were under a “no consumption” advisory based on
levels of DDT and PBB. Total chlordane concentrations in fish from 3 sites exceeded the
trigger level. Thus, even though, according to the state fishing advisories, overall levels
of OC contaminants and mercury are decreasing in the Great Lakes, there is still cause for

concern for Michigan residents consuming sport-caught fish.

2) Mercury

50

Methyl mercury is easily absorbed, rapidly accumulates in most aquatic organisms and
attains it highest concentrations in older, predator fish at the top of the food chain (164).
Unlike the OCs, mercury accumulates in muscle tissue and is thus found at significant
levels in many more fish species. About 90% of the mercury found in fish is
methylmercury (165). US freshwater fish frequently have mercury concentrations
exceeding 1.0 ppm (157). Pike, bass and walleye from contaminated waters tend to have

the highest levels (166).

According to the 1998 Michigan Fish Contaminant Monitoring Program Annual Report
(167), mercury concentrations exceeded the “restrict” consumption trigger level of
0.5ppm in fish from 54% of the sites monitored and the “no consumption” trigger level in
fish from 4% of the sites. Seventy-five percent of inland lakes and or reservoirs sampled

in 1997 had one or more fish with mercury concentrations greater than 0.5 ppm.

c. OC and mercury levels in humans
1) OCs
Human populations resident in the Great Lakes region are at higher risk of exposure
to OCs as a result of the historic contamination of much of the Great Lakes basin with
these synthetic compounds. OC resist degradation, are lipophilic and tend to

bioaccumulate. Most humans have low ambient levels of some or all of these OCs.

According to a Canadian Government Report, 80-90% of human exposure to PCBs and

related environmental contaminants in the Great Lakes region comes mainly from sport-

51

caught Great Lakes fish (168, 169). Great Lakes fish consumers, such as anglers and
fishing boat captains, have increased body burdens of PCBs and DDE compared to non-
fish consumers (170-173), although the levels have declined (145, 174). The number of
exposed people is not insignificant: approximately 2 million Michigan residents and
334,000 nonresidents fish in Michigan each year (175). Both men and women consume
some Great Lakes fish during most of their reproductive years with men consuming more
than women. A variety of species are caught and consumed, often with little regard for

the fishing advisories (176).

PCB levels detected in human serum or plasma from non-fisheaters and fisheaters in the
Great Lakes region have declined in concert with the levels detected in the water bodies
(177). Median PCB levels decreased from 21.4 ppb in fisheaters and 6.6 ppb in non-
fisheaters in from Lake Michigan in 1973-1774 (178) to 5.3 ppb in high fish consumers
and 2.7 ppb in low fish consumers from Montreal in 1994 (Kosatsky et a1. 1996, as
reported in Kearney et a1. ( 179)). In a time trend analysis of levels of Arochlor 1260 in
the serum of a cohort of Michigan anglers, He et al (177) found that the median PCB
levels in male fisheaters changed little from 17.0 ppb in 1973-1974 to 22.9 ppb in 1979-
1982 to 21.1 ppb in 1989-1993. For non-fisheaters, however, the levels decreased from

10.0 ppb to 8.1 ppb to 6.8 ppb in the same time period.
Of the 209 theoretically possible PCB congeners, only a portion are considered to be

environmentally relevant. Many have never been reported in environmental samples, are

not toxic or have low bioavailability (129). In a recent report 89 individual congeners

52

were detected in serum from Michigan fish consumers (145); (Wirth, unpublished data).
However, a subset of these congeners consisting of 25 congeners accounted for over 90%

of the total PCBs found in the serum of Michigan fish consumers (145).

Using data from 59 reports, McFarland and Clark (129) divided the PCB congeners into
four groups based on potential for toxicity (inferred from their ability to induce the CYP
enzyme activities), frequency of occurrence in the environment, and relative abundance
in animal tissues. Based on this scheme congeners 77, 105, 118,126, 128, 138, 156, 169
and 170 were in the group with the highest toxic potential. All but congener 77 are
members of the penta-, hexa, and hepta—chlorobiphenyl isomer groups. The lesser
chlorinated congeners are readily taken up by organisms but are also readily eliminated.
Those more heavily chlorinated congeners (those having seven to ten chlorines) occur in
low concentrations in the environment, remain tightly bound to soil and sediment and are
thus less bioavailable. In the recent re-analysis of PCB congeners found in serum from
members of the original Michigan fisheaters cohort (first assessed in 1979-1982),
Humphrey et al.( 145) focused their analysis on 25 congeners that represented over 90%
of the total PCBs in the subjects’ serum. This list included only three (118, 138 and 170)
from McFarland and Clark’s list of the most potentially toxic congeners. Three congeners
(126,138 and 156) from McFarland and Clark’s list were not found in any subject’s

serum sample.

53

Overall, PCB levels in humans seem to be declining with median levels in the general US
population ranging from 2 to 9 ppb (180). However, selected populations such as anglers

are still at risk through consumption of contaminated sport-fish from the Great Lakes.

2) Mercury
The major source of methyl mercury for humans is fish and shellfish (158) ( 181, 182)
from mercury that has deposited into the Great Lakes and into many of the freshwater
inland lakes. When ingested by humans, nearly 99% of methyl mercury is absorbed,
spreads throughout the body, and is excreted slowly over time (183). The highest levels
are found in the kidney, liver and brain after a few days of exposure. The average person
excretes methylmercury with a half life of approximately 68 days (Institute of Medicine,

1991).

Levels of methylmercury are usually measured in hair, blood, or urine. Little
methylmercury is found in urine, but can usually be detected in hair and blood. The
differential between hair and blood levels varies in different studies (184) (185), but can
be as high as 250 times higher in hair than blood. Concentrations of <20 ppb in blood and
<6ppm in hair are considered to be in the normal acceptable range for adults (186).
Chronic exposure to methylmercury at 0.7 tol .1 ug/kg/day has been estimated to
correspond to hair mercury levels ranging from 10 to 70 ug/g (ppm) and blood level of 44
pg/L (ppm) (183) (in (157)). The upper part of this range may lead to toxic effects in

adults.

54

Health risks of methymercury exposure have been found to occur at varying
concentrations. Exposure of pregnant women to 10 to 20 ppm, as measured in maternal
hair, were associated with motor retardation and CNS toxicity in the child, while levels
for 60 ppm to 500 ppm were associated with paresthesia, ataxia and deafness in adults
(184). Consumption of fish containing 1 ppm, if consumed in sufficient quantity, could
potentially cause levels of methylmercury that have been associated with such adverse
health effects. For example, a hair mercury level of 1 ppm, which can be achieved by
daily consumption of about 0.402 of fish containing 0.5ppm methylmercury, would
produce a dose of 0.1 ug/kg/day, which is the USEPA’s reference dose for causing an
adverse health effect. Reference standards, however vary between regulatory
organizations and range between 0.05-0.07 ug/kg/day to 0.5 ug/kg/day. Additionally, an
adverse health effect would also depend on the particular effect and on individual
variables. WHO has set hair levels of mercury in hair at 6 ppm as the acceptable level and
reports that the estimated average intake of methylmercury is 2.41 rig/day, coming almost

exclusively from fish consumption (186).

3. Effects of OCs on male reproductive outcomes
a. Observations on wildlife and their exposures
Wildlife studies provided the first support for adverse effects of OC exposure on male
reproductive parameters. (187-192). Many animal studies (fish, birds, reptiles, mammals)
have found associations between OC compounds and reproduction (187). Juvenile male
alligator living in Lake Apopka, a Florida lake contaminated with waste from agricultural

activities, a sewage treatment facility and an extensive spill containing dicofol, DDT and

55

sulfuric acid in 1980, exhibited abnormal gonadal morphology, plasma sex steroid
concentrations and greatly reduced plasma testosterone concentrations compared to
alligators from a control lake (Lake Woodruff) (193). Further work indicated that
unstimulated testes from Lake Apopka males synthesized significantly more estradiol
than testes from Lake Woodruff males (5.3 pg/mg tissue/hr versUS 1.1 pg/mg tissue/hr)
(194) while testosterone syntheses was comparable. These results suggest that xenobiotic
chemicals including DDT, modified gonadal steroidogenic activity and hepatic

metabolism of steroids.

The population of Florida panthers has been decimated in part through exposure to a
variety of environmental contaminants found in raccoons, a major source of food.
Necropsies of dead panthers showed potentially toxic concentrations of mercury (l 10
ppm in liver), as well as p,p’DDE, PCBs and other pesticides (195). Additionally, sperm
abnormalities of the Florida panther were the highest reported for any feline and included
low ejaculate volume, low sperm concentrations (3-15x106 spenn/ml), poor motility, and
a very high proportion of abnormal sperm (92.2%) (196). Eleven of l7-free-ranging

panthers and two captive males were cryptorchid (197).

A variety of reproductive abnormalities have been found in wildlife in the Great lakes
region. The increased incidence of teratogenesis and embryonic mortality of colonial
fish-eating water birds from the Great Lakes is at least in part due to exposure to PCBs
(188). Adult bald eagles that have migrated to the Great Lakes shoreline exhibit

reproductive impairment after feeding on fish and other

56

foods from the lakes for 2 or more years (reviewed in Guillette et al (198).

More powerful evidence comes from laboratory studies in which various animals were
fed fish from the Great Lakes contaminated with PCBs as well as other chemicals.
Ingestion of levels of PCBs found in Great Lakes coho salmon was associated with
reproductive failure in mink (199). Similarly, mink fed Great Lakes carp, sucker, perch
scraps, or Whitefish had decreased reproductivity (200). In particular, those fed carp
failed to reproduce at all. It has been suggested that PCBs diminish reproductive
performance in terms of decreased fertility and litter sizes in most animal species tested

under laboratory conditions (201 ).

b. Effects on spermatogenesis
Early studies focused on treatment of adult experimental animals with PCBs, usually as
Arochlor mixtures, and TCDD, found increases in testes weights (202) (203), reductions
in sperm counts (204), decreased fertility (Khera and Ruddick, 1973) and altered plasma
androgen levels (205) (206). However, the doses required to achieve these effects were
often toxic and far above levels ever seen in human exposures. For example, a single
dose of PCB congener 77, which has a high TEQ, of 18mg/kg (ppm) resulted in
permanently reduced daily sperm production and a significant increase of abnormal
sperm (207). For comparison, the median serum level of men exposed to PCBs through
fish consumption in the 1970’s was 21.4 ppb (208). Similarly, for TCDD, the ED50 (the
dose that will cause an effect in 50% of the animals), for androgenic deficiency is 15,000

n g/kg (205), while the average background body burden for humans is 8-13 ng/kg (209).

57

Subsequently, investigators shifted focus to more sensitive periods in mammalian
development to determine if exposure to contaminants in the range of human exposures
during these periods would have an effect on reproductive outcomes. Animals exposed in
utero and perinatally, when organogenesis and further organ development, respectively,
occurs, required significantly lower does of CC to produce adverse reproductive effects
(reviewed in (210). Exposure of experimental animals in utero appears to be the period of

development most sensitive to the harmful effects of DC (142).

More recent data indicate that for both experimental animals and humans, puberty is also
a time when lower exposures can lead to adverse reproductive effects (21 1). During
puberty, levels of gonadotroping-releasing hormone (GnR) produced by the
hypothalamus increases, which leads to increased levels of gonadal sex hormones and
secondary sexual development (212). In males secretion of FSH leads to the development
of seminiferous tubules and an increase in testis size. LH release leads to testosterone
secretion by the Leydig cells while FSH secretion leads to production of inhibin secretion
by Sertoli cells, thus establishing the regulatory networks in the hypothalmus-anterior
pituitary—gonadal axis. Exposure to hormone disrupting chemicals, such as PCBs and
TCDD, duringthis time may have adverse effects on spermatogenesis. Since these
chemical are lipophilic and have long half-lives, the effects could persist into adulthood

(211).

58

Since the focus of this proposal is exposures occurring from adolescence through

adulthood, this section will review studies covering these life periods.

1) Effects of exposure to PCBs on experimental animals and
humans
a) Experimental animals
Studies on experimental animals exposed to PCBs have added increased support for the
notion that exposure to PCBs explains some of the effects of OC exposure on wildlife.
PCB treatment of adult rats with a single subcutaneous dose of 18mg/kg or 60 mg/kg
body weight of 3,3’,4,4’- tetrachlorobiphenyl (congener 77 which has a TEF value of 10‘
3 (132) resulted in a reduction of daily sperm production from the control value of 31x106
sperm to 22x 106 and 20x 106 sperm, respectively, for the two PCB groups at 8 weeks.
The sperm number per caudal epididymus was significantly reduced one week after
treatment from the control level of 222 x106 sperm to 135 x 106 and 142 x 106 for the
PCB treatment groups. Significant increases in the percentage of abnormal sperm were
observed in both treatment groups compared to the control group, with changes in the

tails being the most frequently observed abnormality (207).

In another study, female mice were fed diets containing 30 ppm daily of PCB congener
77 for two weeks before pairing to untreated males. Females received the same daily
dose throughout mating, gestation and lactation. F1 males were given the same diet
postnatally (213). At three weeks of age, PCB-exposed male mice had significantly

increased testes weight compared to unexposed controls (43.2 mg vs. 32.9 mg,

59

respectively, (p<0.05). The percentage of eggs fertilized in vitro by sperm from similarly

PCB-treated mice at 19 weeks of age was significantly reduced to 46.7% compared to

57.9% fertilized by control mice (p<0.05).

In a third study, the effects of exposure of male rat pups to Arochlor 1254 during
lactation on their adult reproductive functions were investigated (202) (214). During
lactation in rats, development is continuing and transfer of PCBs is greater than during
gestation (215). Male pups were exposed to PCBs in the milk of mothers given Arochlor
1254 on days 1,3,5,7, and 9 day of lactation, at doses of 8mg/kg, 32mg/kg and 64mg/kg
(214). Beginning on 130 days of age, the PCB-exposed males were mated to unexposed
females. The number of embryo implants, the number of embryos and the implants per
corpora lutea were all significantly reduced in females mated to males exposed to
32mg/kg and 64mg/kg Arochlor compared to females mated to unexposed males
(p<0.001). In a follow-up study by the same authors using a similar protocol and
Arochlor doses, a significant decrease was observed in the number and percent of normal
fertilized eggs and eggs at the two-to four blastocyst stages in females mated to Arochlor-
ex posed males compared to females mated to unexposed males (216). No reduction in
caudal sperm number, sperm production, epididymal sperm morphology and either FSH
or testosterone levels was observed. While there are significant differences in the
protocols used by these three groups (PCB congener(s) type, dose, and route), they
indicate that exposure of animals early in life to PCB congener mixes or specific PCB

con geners can lead to adverse effects on fertility and sperm function during adulthood.

60

b) Effects of PCBs on humans
Effects on human sperm parameters have been harder to document, in part because of the
difficulty of obtaining semen samples. A study compared the levels of 74 PCB
congeners in 33 fertile men, 50 subfertile men (20-60 x 106 sperm/m1), 50 men with
idiopathic oligosperrnia (<20 x 106 sperm/ml) and 25 men post vasectomy (N=170). A
significant correlation was noted between sperm motility and the levels of congeners
123, 138, and 167 in samples with sperm counts below 20 x 106 in regression analysis
(217). The mean total PCB level in seminal fluid was 5.8 ppb and the maximum levels of
the three congener were 2.3 ppb (congener 123), 1.1 ppb (congener 138) and 1.1 ppb
(con gener 167). In a more recent pilot case-control study investigating the relationship
between OC and human sperm parameters, Hauser et al. (218) showed that men with
decreased sperm counts (<20 x 106/ ml), decreased sperm motility (<50%) and decreased
percentage of morphologically normal sperm (<40%) had higher levels of PCB congeners
1 18, 138 and 153 than controls with normal semen parameters. Due to their small sample
size (29 men), these results can be no more than suggestive. Recently, a modest
association in men with high levels of Lake Michigan sport-caught fish consumption and
risk of conception delay (OR=2.8 CI, 1.0-8.0) has been found (219). Another study
conducted with sport fish consumers from Lake Ontario, however, did not find such an
association (220). Both studies obtained the analyzed data cross—sectionally and both had
low participation rates for the over all study: 29% (219) and 39% (220). Each study used
a different method to calculate the number of fish meals consumed, neither of which was
precise, so it is difficult to compare the levels of exposure. Additionally, differences in

fish species, location, and contaminant types and concentrations between the two studies

61

may contribute to the differing results. A study of transformer repair workers did not find
an association between level of PCB exposure and sperm counts (221). However, the
levels of individual congeners was not determined and PCBs were not analyzed as a
continuous variable, which may have decreased their ability to detect smaller differences.
Other studies have also detected PCBs in human semen with mean values for total PCBs
of 5.8 ng/g (217), 7.0 ng/g (222) and 1 1.7 ng/g (223), but did not correlate these levels

with sperm parameters.

A follow up study of individuals exposed to PCBs and dibenzofurans in the large scale
poisoning in Taiwan in 1979 through ingestion of contaminated rice oil (Yu-Cheng)
found an effect on male children exposed in utero (224). All prenatally exposed men 16
years or older and 48 healthy volunteers of comparable age and no history of chemical
exposures were recruited. A significant difference in the percentage of morphologically
abnormal sperm (37.5% in 12 exposed men and 25.9% in 23 unexposed men, p<0.0001)
was found. A significant reduction in the percentage of total motile sperm (35.1% for
exposed men, 57.1% for unexposed men, p<0.0058) was reported. Additionally, a
reduction in the percentage of hamster oocytes penetrated (65.8, exposed and 73.5,
unexposed (p<0.017) was found. No differences in semen volume or sperm count were

found.

In summary, while experimental animal studies indicate adverse effects of PCB on male

reproduction and on sperm parameters, data from human studies are sparse. It is

62

noteworthy, however, that those studies in which levels of individual PCB congeners

were measured found significant reductions in sperm function.

2) Effects of OC pesticides on semen quality
1) Dibromochloropropane (DBCP)

OC pesticide exposure has also been shown to affect spermatogenesis in several studies.
Similar effects of DBCP on spermatogeneis were reported for production workers in
pesticide factories using DBCP as an ingredient in California (225) (226) and Israel (227-
229) in 1976 and 1977, respectively. Direct quantitation of DBCP exposure levels were
not made in either study, so length of time at their job was used as a proxy for exposure.
In investigations of both exposures, fertility problems were reported only by men coming
into direct contact with DBCP during its production. In both studies, semen analysis of
DBCP exposed men revealed decreasing numbers of sperm with increasing years of

exposure.

The California study began with the 39 people working in the Agricultural Chemical
Division of the pesticide plant, but after excluding women, none of whom complained of
fertility problems, men with vasectomies or men with intermediate sperm counts (10-30
x 106/m1), ended up with 11 men with normal (> 40 x 106 sperm/ml; group A) and 1 1
men with low (<1 x 106 per/ml; Group B) sperm counts. Nine volunteers consented to a
testicular biopsy. Significant differences between Groups A and B (p<0.01) were found
for years exposed (A, 8.0 y vs B, 0.08 y), sperm count (A, 0.2 x 106/m1 vs B, 93 x 106/m1

), FSH (A,11.3 vs B, 2.6 mi.U/ml), and LH (A, 28.4 vs B, 14.0 mi.U/ml ) but not with

63

testosterone (459 vs B, 463 ng/dl). Two men in Group B who were not azoospermic (0 x
106/m1 sperm) had sperm with decreased motility and increased morphological
abnormalities. Biopsies of men in Group B revealed loss of spermatognia. A high level
of DBCP exposure in the current year was significantly associated with depressed sperm
counts (p<0.01)(230). The level of exposure to DBCP in early 1977 was measured at 0.4

ppm USing personal air monitoring devices.

The Israeli study found that longer, continuous exposures to DBCP by production
workers was associated with a decrease in sperm count and, if long enough, to
oligospermia (greater than 0 but less than 20 x 106 sperm/ml), azoosperrnia and sterility
(231). Testicular biopsies on an azoospermic man showed severe atrophy of the germinal
epithelium with no evidence of active spermatogenesis (107). Follow up of the 30
exposed men at five years (231) found that some of the oligospermic and azoospermic
men had recovered and fathered children. Azoospermic men who did not recover had
FSH levels that were almost three times higher than in men who recovered
(approximately 9 leml versus 25 mIU/ml, respectively). LH levels were initially
comparable in both groups of azoospermic men, but by four years the levels increased
and stayed elevated in men who did not recover (approximately 10 mIU/ml versus 19

mIU/ml, respectively). Testosterone levels remained normal in all men at all times.

A study conducted in 1977 of California pesticide workers exposed to DBCP included a
total of 96 men from 5 counties who came in for examination (230). Ten men reported

clinical infertility, and two had sperm counts less than 1 x 106 sperm/ml. An association

between mean sperm count and current year exposure (61.7 x 106 spemi/ml for zero days
to 21.9 x 106 sperm/ml for greater than two months) was significant (p<0.01), while the
trend for past years exposure to DBCP (61.7 x 106 sperm/ml for 0 years to 30.2 x 106
sperm/ml for eight or more years) was marginally non significant for trend (p<0.054). A
significant association between FSH levels and days exposed in current year, but not LH
levels, was significant (p<0.5). They also suggested that the effects of DBCP appeared to
be reversible, since lower sperm counts were associated with current rather than past
years exposure. Thus DBCP, a halogenated hydrocarbon, has been shown to have toxic
effects on the spermatogonia, which depending on the exposure, can lead to long-term
sterility. While DBCP may seem like an extreme example, it should be kept in mind that
over 1000 chemicals have been identified in the waters of the Great Lakes and only a
fraction of them have been tested for any type of adverse health-related effects, much less
for their ability to interact with each other.

b) Other pesticides
Occupational exposure to pesticides has been reported to affect time to pregnancy, a
measure of fecundability. A correlation between pesticide spraying season and time to
pregnancy was found for fruit growers in the Netherlands exposed to pesticides (de cock
et a1. 1994). On the other hand, Savitz and associates assessing the potential reproductive
hazards associated with farming in the Ontario Farm Family Study, were not able to find
significant associations between farm chemicals and time to pregnancy (232) or an
altered sex ratio (233). Similarly, no association was found between farmers and
agricultural workers in France and Denmark and time to pregnancy (234). Differences in

study populations and chemicals examined may explain these contrasting results.

65

Pesticides have been shown to disrupt endocrine function (235). P,p’DDE
(dich1orodiphenyldichloroethene), the major and persistent metabolite of DDT, was
shown to be a potent androgen receptor antagonist capable of inhibiting the effects of
androgens (236). Effective doses were in the range of those found in humans in areas
contaminated with DDT. Of interest, the biologically active metabolite of methoxychlor
(the methylated isomer of DDT and a xenoestrogen), was shown to significantly reduce
testosterone production by Leydig cells in vitro (237). This effect was mediated via a
decrease in the activity of CYP] 1A1, a P450 enzyme involved in steroidogenesis.
O,p’DDT (an isomer of DDT) as well as p,p’DDE, and p,p’DDT have been found in

human semen (238).

3) Effects of dioxin
Much of the evidence supporting a role for dioxins and male fertility has come from
experiments with animals. Dioxins can affect male fertility in experimental animals if
administered to adult animals although their strongest effects seem to occur after in utero
exposure. In post pubertal experimental animals, dioxin, at doses causing overt toxicity,
causes decreases in the weight of the testis and the accessory sex organs, decrease in
spermatogenesis, abnormal testicular morphology, inhibition of testicular steroidogenesis,
reduction in plasma androgen concentrations and adverse reproductive performance (239-
244). If administered during gestation at a time when the hypothalamic/pituitary/testis
axis is developing, a single dose of dioxin adversely affects the male offspring by
decreasing serum androgen concentrations and impairing androgen-dependent perinatal

development (245), by caUSing demasculinization and feminization, by impairing sexual

66

differentiation of the central nervoUS system (246), and by decreasing spermatogenesis
(247). Maternal and lactational exposure of rats to dioxin increased the frequency of
cryptorchidism in the male offspring, caUSed dose-dependent reductions in male
offspring of testicular weight (247) and sperm counts (248). Sperm from male rats
exposed to dioxin in utero had decreased motility, decreased in vitro fertilizing capacity
(213) and permanently altered sperm-transit through the epididmeS (210, 249). Dioxins
thUS appear to have clear cut effects on spermatogenesis and male sexual development in
animals but at doses much higher than seen in humans. However, if the exposure occurs
during sensitive developmental times, such as during puberty, they may potentially affect
spermatogenesis at doses seen in human studies (21 1). A target of these toxic effects is
the Leydi g cell, whose number and steroidogenic enzyme activity are reduced after

TCDD treatment of adult rats (205).

Male workers exposed to high levels of dioxin where found to have statistically
significant, elevated levels of LH and FSH, and decreased levels of testosterone (250). A
study on the risk of infertility and delayed conception and workplace exposures found
that fishermen had a significantly increased risk of sperm abnormalities (p<0.05)
although the specific exposure was not determined (251). Dioxins have been detected in

human semen (252).
In summary, the evidence suggests that PCBs, pesticides and dioxin can adversely affect

various aspects of male reproduction in animals and in humans if the dose is high enough

or if the exposure occurs during a susceptible time during development. From the FFHP

67

data we know that 80-90 % of male participants reported eating sport-caught Great
Lakes fish during their teens and twenties, possibly times when the effects of these
contaminants may be harmful to processes involved in spermatogenesis and regulation

and function of reproductive hormones.

b. On the sex ratio
The male proportion, commonly referred to as the sex ratio, is number of live males
births divided by the total number of live births. For most western countries this value is
usually 0.512 to 0.515 and remains stable (114) (1 15). The ratio is determined early in
gestation when sexual differentiation takes place. Prenatal sexual development is a
complex process requiring balance and timing of exposure to androgens and estrogens
(reviewed in (253) (115)). Initially all embryos appear to be female and will remain
phenotypically female without androgenic stimulation. Between 6 and 9 weeks of
gestation, Sertoli cells of the testis or follicular cells of the ovary become active. At this
point the embryo has a unisex pair of gonads and two sets of ducts, referred to as the
wolffian and mullerian ducts. In genetic males, the gonads differentiate into testes and
produce testosterone, which further drives masculine development. Testosterone
stimulates the differentiation of the wolffian ducts into the ductus deferens and the
descent of the testes into the scrotal sac. Mullerian inhibitory hormone, produced by the
Sertoli cells, directs the regression of the mullerian ducts. In genetic females, the wolffian
ducts disappear in the absence of androgens and the mullerian ducts develop into the

oviducts. Normal differentiation of the testes depends on the undisrupted functioning of

68

the Sertoli cells at this critical stage. Thus, exogenous agents capable of changing
androgen or estrogen levels (endocrine disruption) or damaging to the cells of the testes
at this time could affect the phenotypic determination of sex and subsequent

development.

It has been hypothesized by James that mammalian parental hormone levels at the time of
conception affect the sex ratios of the young at birth (James 1996). Thus the sex ratio of
offspring is increased by high levels of testosterone and estrogen in the parents at the

time of conception, and decreased by high levels of gonadotrophin and progesterone.
Exposure to environmental contaminants with the potential to modulate gonadotrophin
levels has been linked to changes in the sex ratio through studies of accidental and

occupational exposures.

1) Effects of PCBs and PCDFs on the sex ratio
Two mass food poisonings of contaminated rice oil, called Yusho in Japan (254) and
Yucheng in Taiwan (255), resulted in widespread exposures to PCBs and PCDFs. In
Taiwan, about 2000 individuals were affected from 1978 to 1979 and the estimated
amount of PCBs ingested was 0.7 to 1.84 g (255). Serum PCB concentrations from 613
individuals ranged from 3 to 1,156 ppb. Recently, the sex ratio of 137 live births in
Taiwan from 1978 to 1985 by 74 women who had registered with the health department
as being affected was determined (256). Sixty-nine girls and 68 boys were born giving a
ratio of 0.496. The authors state that the number of girls born was not excessive, but do
not provide the sex ratio of a comparison group, such as children born to unexposed

mothers. For comparison, the sex ratio for Japan for 1978-79 was 0.514 (124). In this

69

study, the eligibility criteria for the women included having at least one child alive at the
time of the interview in 1985, which could be up to 8 years after the exposure. The reason
for this criterion is unclear, since if anything, it would bias the ratio in favor of more girls
since boys die at a faster rate than girls at all life stages. Additionally, there is no
information on the exposure levels of the 74 women. This could bias the results in either
direction since these volunteers could either be more health conscious and perhaps
healthier or they could be the most seriously affected. No consideration was given to the
exposure status of the fathers, which again could have increased, if exposed, or
decreased, if unexposed, the effects on the sex ratio. Thus the effect of PCB exposure on
the sex ratio cannot clearly be ascertained from this report. No reports on the sex ratio of
children of the Yusho patients could be found to evaluate the effects of PCBs and

PCDF’s on the sex ratio for that population.

Experimental animal studies provide some additional evidence for the effect of PCBs on
the sex ratio, although the exposures were in utero. In a study on the effects of Arochlor
1254 on the sex ratio of rhesus monkeys, 80 menstruating rhesus monkeys were untreated
or treated with from 5 to 80ug Arochlor 1254/kg/day for 25 months. A total of 36
impregnations occurred (257). Of these, the sex could be determined on 26 live births,
stillbirths and abortions. The sex ratio for the offspring of the untreated monkeys was
0.500 and for the treated, 0.4375. Of the abortions and stillbirths in the treated group
whose sex could be determined, all were male. The authors conclude that there was some
suggestion that Arochlor 1254 may adversely affect the viability of the male rhesUS

fetUS. In another study, pregnant female rats were dosed with 6 mg/kg of 3,3’,4,4’

70

tetrachlorobiphenyl (PCB 77) daily by gavage from day 6 to day 18 of gestation (258).
PCB 77 reduced the sex ratio before and after birth from 1.00 (control) to 0.72 and 0.69,
respectively. While both sexes were affected, males were preferentially killed. Significant
mortality occurred late in gestation, on days 19 and 20, suggesting that the congener

acted directly on the fetUS rather than by affecting maternal physiology.

b) Effects of dioxin exposure on the sex ratio
Studies on a human dioxin exposure resulting from an explosion in a herbicide plant in
1976 in Seveso, Italy, assessed the effects on the sex ratio of children born to exposed
parents (259). Seventy-four children born in the high dioxin exposure zone (A-zone)
from 1977 to 1984 had a sex ratio of 0.351 (26 males vs 48 females), compared to the
expected ratio of 0.514 ( X2 test, p<0.001)(260). In families in which both parents were
in the A-zone at the time of the explosion and also had high serum dioxin levels (104-
2340 ppt in 1976), no males were born from 1977 to 1984. In a more extensive analysis
of the sex ratio of children born to exposed parents in whom serum dioxin levels were
determined, it was found that exposure of both parents to greater than 15 ppt dioxin
decreased the sex ratio to 0.442 (p=0.03, compared to the expected ratio of 0.514)(21 1).
Only the father’s exposure was shown to significantly decrease the sex ratio and the ratio
decreased with increasing dioxin concentrations (X2 for trend, p=0.008). Fathers exposed
before or during puberty (younger than 19 years old) had a lower sex ratio than those
exposed later in life (0.382 versus 0.469, respectively), suggesting that the time before
and during puberty may be a very sensitive period for dioxin toxicity in men.

Additionally, fathers who had greater than 15 ppt serum dioxin levels and who were

71

exposed during puberty continued to father significantly more girls than boys more than
15 years after the 1976 exposure, suggesting that these effects of dioxin may be long-

lived, if not permanent.

Thus, exposure of males to environmental chemicals during critical periods of their
development can affect aspects of their fertility involved in the determination of the sex
ratio. It has in fact been suggested that the timing of such exposures may be more critical

than the total dose rate in determining a broad range of outcomes (261).

Servicemen in the Vietnam War were also exposed to dioxin as a result of extensive
aerial spraying with the herbicide Agent Orange. An ingredient of Agent Orange, 2,4,5-T,
was contaminated with dioxins at a mean concentration of 2 ppm (262). Exposure
assessment was based on service location as determined from self-administered
questionnaires and service records or spraying locations (263). Exposure to Agent
Orange was not associated with any of the reproductive outcomes measured including
difficulty in conception, time to conception or the sex ratio of the offspring. Since the
risks of exposure to Agent Orange were unforeseen, semen quality was not assessed at
any time following exposure. The study design was ecological and as such individual
exposure levels were not determined. Additionally it is likely that assessment of location
by self-report and by official records may not accurately reflect the actual war-time
situation. It thus difficult to determine the exposure parameters and correctly assess the
effects of Agent Orange on specific individuals and their reproductive health from this

study.

72

3) Effects of OC pesticides on the sex ratio
a) Dibromochloropropane (DBCP)

OC pesticide exposure has also been shown to affect the sex ratio in several studies. The
study investigating the effects of DBCP exposure in pesticide workers in Israel described
above also found alterations in the sex ratio. Follow up of the 30 exposed men in the
Israeli study at five years (231) found that some of the oligospermic and azoospermic
men had recovered and fathered children. Of the 13 children born to exposed fathers, 11
were girls, yielding a sex ratio of 0.154 (p=0.011)(228, 229). The sex ratio of men who
were normosperrnic (greater than 20 x 106 sperm/ml) was 0.80 (5 live births) compared to
0.53 for the pre- or unexposed men (p>0.05). The sex ratio of men who remained
oligospermic was 0.25 while that of men who remained azoospermic was 0.00 (4 live

born females) (p< 0.015 for oligospermic and azoospermic men).

The Israeli group of men were evaluated again at 8 and 17 years post exposure. At 8
years, 15 formerly azoospermic men and 7 formerly oligospermic were followed up and
their sex ratio compared with the sex ratio of men formerly unexposed and men with pre-
exposure live births. The sex ratio of the pre-exposed and unexposed men (a total of 51
live births) was 0.529. The sex ratio for pregnancies occurring during the exposure (17
live births), 0.352; and for pregnancies occurring during the recovery period (19 live
births), 0.210 ( 107). At the 17 year follow-up of 9 formerly azoospermic and 6 formerly
oligospermic men, no further recoveries had occurred. Of the 41 live births of the 7

recovered men, 17 were male, yielding a sex ratio of 0.414 (264). Thus the effects DBCP

73

exposure on the sex ratio was severe and long lasting, much like that seen with the dioxin

exposure in Seveso.

Another study did not find that exposure to DBCP affected the sex ratio. They found that
exposure to DBCP in drinking waters did not adversely affect birth outcomes, including
the sex ratio (265), even at the highest level of contamination (1-3 ppb in the drinking
water). This was an ecological study, however, with exposure being determined by
census tract residence rather than individual body burdens. As such, a direct association

between DBCP exposure and the sex ratio cannot be made.

2)) Clordecone (kepone)
Clordecone, a chlorinated hydrocarbon insecticide, was involved in one of the most
costly chemical disasters in the United States (266). The disaster occurred in a small,
single-product manufacturing factory, named Life Sciences Products Company, which
made the insecticide Kepone for Allied Chemical Corporation in Hopewell, Virginia. In
1974 and 1975, 76 of 133 people working at the plant developed a clinical illness
associated with production work with kepone. Kepone blood levels in workers with the
illness were 2.53 ppm and those without the illness were 0.060 ppm (p<0.001) (267).
Thirteen patients (17.1%) had oligospermia with abnormal and nonmotile forms

predominating (268).

Taken together, these results suggest a link between damage to the germinal epithelium

by DBCP, decreased sperm counts and decreased sex ratio. These results suggest a

74

continuum of deleterious effects on the spermatogonium by increasing levels of exposure
to environmental toxicants like DBCP and dioxin. On a population level, the first effect
may be a decrease in the sex ratio followed by falling sperm counts and ultimately by
increases in sterility and/or testicular cancer. Movement from the first effect to the
subsequent ones may involve cumulative exposures of related chemicals, which may
interact to amply their effects as well as exposure to very high doses of a single chemical,
which rarely occurs. For example the effects of both DBCP and dioxin in Seveso (260)
appear to be reversible up to a certain level of exposure, after which the higher doses may

cause permanent damage to the germinal epithelium.

(I. Comparison of animal and human OC doses

DeVito et al. (209) compared estimated body burdens of dioxin-like chemicals reported
in studies on humans and experimentally animals associated with the same health
outcome. In humans, body burdens were estimated from lipid-adjusted serum
concentrations of dioxin-like chemicals, including dibenzo-p-dioxins, dibenzofurans and
PCBs using the TEF method. The value obtained thus reflected the body burden of
chemicals with dioxin-like properties and was expressed as the TCDD equivalent factor
(TEF). In the general human population, average background concentrations were
estimated as 58 ng TCDD equivalents (TEF)/kg serum lipid, which corresponds to a body
burden of 13 ng TEQ/kg body weight. Human populations with known exposures to

dioxins have body burdens of 96-7,000 ng TEQ/kg body weight, or about 8 to 538 times

75

the average exposure level. Background levels of TCDD were 1 and 4 ng/kg in rats and

mice, respectively.

For effects clearly associated with dioxin exposure, such as chloracne and induction of
CYP1A1, a gene induced by TCDD-like chemicals that regulates hormone and
contaminant metabolism (see section 4. for more details), humans and animals responded
at similar body burdens. Development of chloracne in humans occurred at body burdens
of 96-3,000 ng/kg while the animal dose was 1,000ng/kg in monkeys and 13,900 ng/kg
(4,000 ng/kg, 3 days/week/2 weeks) in mice. The lower human exposure value, 96 ng/kg,
was found in an individual with the lowest reported adipose dioxin concentration for any
human with chloracne. This individual was exposed to a mixture of chlorinated
dibenzodioxins and chlorinated dibenzofurans. The higher value, 3,000 ng/kg, represents

the average body burden of TEQs in individuals from Yusho with chloracne.

Induction of CYP1A1 in human placenta occurred at 2,130 ng/kg while induction in rat
liver required 2,582 ng/kg (125 ng/kg/day, 5 days/week/l3 weeks). The human exposure
level was obtained in placentas from mothers highly exposed during the Yu-Cheng
incident. These mothers gave birth to babies with lower birth weights compared to
babies from unexposed mothers and had altered levels of placental CYP1A1 and
placental epidermal growth factor receptors. Induction of CYP1A1 activity was also
measured in German chemical plant workers exposed to TCDD (Masten et a1 1997).
Elevated levels of CYP1A1 enzyme activity were found in 34 workers diagnosed with

chloracne with TEQ values ranging from 22.7 to 914.7 ppt.

76

In contrast, the body burdens associated with the less clearly defined carcinogenic effects
of TCDD ranged from 944 ng/kg in mice to 137,000 ng/kg in hamsters. In
epidemiological studies that reported an association between TCDD exposure and cancer,
body burdens were estimated between 109 and 7,000 ng/kg at the time of highest human
exposure. For example, a report on cancer incidence from 1977 to 1986 in individuals
exposed to TCDD in Seveso, Italy, in 1976, found significant increases in cancers of the
hematopoietic system in men and hepatobilliary system in women (Bertazzi et a1 1993),
although overall cancer rates were not increased. The increases occurred in residents of
Zone B, whose blood concentrations of TCDD were estimated at 74 to 526 ppt (ng/kg).
No increases in cancer incidence were found in the 724 residents of Zone A, perhaps
because, at the time of this publication, there were only 14 reported cancer cases. Thus,
the TCDD levels associated with cancer development in this group are approximately 5
to 40 times the background level for humans, but are much lower than the dose required

to cause cancer in rodents.

In contrast to the better established effects of dioxin-like chemicals that occur at
relatively high exposure levels, several other effects occurring at variable doses have
been reported in experimental animals. Decreased sperm counts in rats occurred at a
TCDD body burden of 64 ng/kg (64 ng/kg, maternal dose) (247), while testis
abnormalities in rats and mice occurred at 12,500 (rats)(205) and 100,000 ng/kg in mice
(240). Sperm from rats exposed during lactation to mothers given a total of

150,000ng/kg Arochlor 1252 during lactation (32ug/kg on days 1,3,5,7 and 9) had

77

significantly decreased linear motility (p<0.5) (216). In humans, significantly decreased

sperm motility occurred with seminal fluid PBC congener concentrations of 1.1 to 2.3

ng/g (ppb) (217).

Little can be concluded on close equivalences from these experiments on reproductive
outcomes since the actual amounts of PCBs to which the males rodents are exposed is not

known.

In conclusion, it appears that for outcomes causally associated with TCDD exposure, the
levels required in experimental animals and humans are similar, and higher than
background levels for both species. For cancer, the levels required to see an effect in
humans is appreciably lower than in animals, but still higher than the levels seen in the
general population. For reproductive comes, the data is much more limited and

insufficient to draw any conclusions.

4. Mechanisms of OCs effects on male reproductive paramters
a. Role of genetic polymorphisms
1) In genes encoding cytochrome P450 enzymes
Enzymes coded by cytochrome P450 genes, also referred to as CYP genes, play critical
roles in the activation and detoxification of a wide variety of environmental toxicants and
in the metabolism of sex hormones, including hydroxylation of estradiol and
testosterone. Some of these genes have stable genetic variants, or polymorphisms, that

encode enzymes with no, increased or decreased activity. Some of the genes are inducible

78

by a variety of chemicals including environmental contaminants. Substrates for the
enzymes also include a wide variety of environmental contaminants, drugs, and steroid

hormones, and the substrates are often the same as the inducers.

Dioxin and the coplanar PCB congeners induce transcription of CYP genes involved in
activation of environmental contaminants to chemical forms able to damage a variety of
human systems (269, 270). CYP1A1 and CYPlBl catalyze the formation of mutagenic
intermediates from several polycyclic aromatic hydrocarbons, some of which are potent
mammary gland carcinogens in rodents (271, 272). Other CYP genes induced by dioxin-
like compounds are involved in the metabolism of estrogen and testosterone (273). The

basic mechanism is described in the following figure (Fig 3).

     

 

/
/
/ <
/ __ x
/ 1”"! ‘ \
// // \‘\?
// / A?
,t’ \|
if I"! \|
1 I 1
1 3, If t§§ CYPIAI gene 1
\ Eve-ifiéom/
‘ \ )an /
\ \\ A
\. "Em” e P4501A1
\ < FR ”3
\\ I
\“K\ /

Figure 3. Model for dioxin activation of genes via AhR binding .
Dioxin (L) binds to AhR/hsp90 complex, releasing hsp90 and creating a binding site for
the AhR nuclear transporter (Amt). This complex translocates to the nucleus and binds to
xenobiotic receptor elements (XRE) in the DNA and initiates transcription of the
CYP1A1 gene. The CYP1A1 RNA is transported to the endoplasmic reticulum (ER)
where the CYP1A1 message is translated into the CYP1A1 enzymes (from Kawajiri and
Hayashi (274).

79

The toxicity of dioxin and the coplanar PCBs is positively correlated with their ability to
bind to the aryl hydrocarbon receptor (AhR) (133) and induce transcription of several
CYP genes, including CYP1A1, CYP1A2 and CYPlBl (269) (273). In a study on the
effects of exposure to dioxin-containing OC herbicides and pesticides, the transcriptional
levels of CYP1A1 in peripheral blood lymphocytes from German chemical workers were
shown to increase with increasing dioxin exposure levels (275). The serum levels of
dioxin in exposed workers with chloracne (n=l6), a skin disorder associated with high
dose dioxin exposure, was 32.1 ppt compared to 11.5 ppt in exposed workers without
chloracne (n=15). The corresponding levels of CYP1A1 activity, expressed as EROD
(ethoxyresorufin O-deethylase) activity, were 2.81 and 2.32 pmol/min/mg, respectively.

CYPl B1 activity is also induced by dioxin (276) and other OC compounds (272).

Both human CYP1A1 and CYPlBl encode enzymes that are also 17B-estradiol (E2)
hydroxylases; CYP1A1 is expressed primarily extra-hepatically and CYPlBl is
expressed primarily in the breast (reviewed in (277). CYP1A1 catalyzes the
hydroxylation of E2 at the C-2, -6a and -15a positions while CYPlBl is more active at
the C-4 position (278). The 4-hydroxylated metabolite is elevated in human malignant,

but not normal breast tissue (279) and is carcinogenic in animals (280).

Several polymorphic CYP genes that are induced by PCBs have been shown to be

associated with increased risks of cancers of the reproductive tract possibly by altering

reproductive hormone metabolism or increasing reproductive toxicant levels.

80

Four polymorphisms have been identified in the human CYP1A1 gene (reviewed in
(281), one of which has been associated with postmenopausal breast cancer and PCB
exposure (282) and 10-15% of white Americans have an allele with this substitution
(283). The mutation does not appear to alter enzymatic activity (284), but the activity
seems to be more inducible than the wild type gene (285). In a case control study of 154
women with postmenopausal breast cancer, women with the CYP1A1*2B polymorphism
(substitution of valine for the wild type isoleucine) in at least one chromosome who were
also exposed to high levels of PCBs [3.73-19.04 ng/g (ppb) of serum] had a significantly
increased risk for postmenopausal breast cancer (OR, 2.96; 95% CI, 1.18-7.45) (282). In
this case-control study there is a possibility that the cachexia associated with cancer
resulted in weight loss and mobilization of PCBs from fat stores to the blood,
artefactually increasing serum PCB levels. This issue was not addressed by the authors.
Additionally, the possible interaction between this CYP1A1 polymorphism and cigarette
smoking, previously shown to significantly increase breast cancer risk (286) (287), was
not investigated due to the small sample size. Thus while the results of this study are
suggestive of an interaction between a genetic variant involved in toxicant and ES
metabolism and a hormone-dependent cancer, the association needs to be evaluated with
the proper controls (women without cancer but with cachexia and the same
polymorphism) or by using a prospective study design. The study does highlight,
however the possible interaction between a PCB-inducible gene and alterations in sex

hormone metabolism.

81

In an earlier report in which potential interactions with DC exposure were not
investigated, the same polymorphism was associated weakly and nonsignificantly with
postmenopausal breast cancer risk (286). In light smokers, however, the presence of at
least one allele with the polymorphism was significantly associated with increased breast
cancer risk (OR, 5.22; CI,1.l6-23.56) although the numbers were small (N=29). No effect
was seen in heavy smokers (20 or more pack-years), perhaps due to the reported anti-
estrogenic effects of smoking (288). Similar results were reported by Ishibe et al.(287)
who found that women who had commenced smoking before the age of 18 and had the
CYP1A1 2* genotype had an increased risk of breast cancer (RR, 3.61; 95% CI, 1.1 l-

l 1.17). The authors interpreted their results to suggest that exposure to the carcinogenic
polyaromatic hydrocarbons found in cigarette smoke induced expression of the mutant
CYP1A1 allele, which codes for an enzyme with an increased capacity for metabolizing
estrogen to a metabolite (l6-hydroxy estrogen) that is an estrogen agonist (289, 290).
Alternatively, since CYP1A1 is also involved in activation of genotoxic substances with
the ability to produce reactive oxygen intermediates capable of damaging DNA (291)
(292), increased inducibility of the mutant gene or production of an enzyme with

increased activity could lead to increased DNA damage.

CYP3A4 is another gene involved in xenobiotic metabolism and in the hydroxylation of
testosterone to a form that is eliminated. The gene has been shown to be induced by
PCBs in monkeys (293). Men with a genetic polymorphism in CYP3A4, CYP3A4-V,
were found to have a significantly increased risk of prostate cancer (294). Significant

trends for all men regardless of family history (p=0.0003) or men with no family history

82

(p=0.0008) for any stage of tumor diagnosis or increasing severity of tumor stage at
diagnosis and the polymorphism compared to men without the polymorphism were
found. The polymorphism was found in the 5’regulatory element of the gene, possibly
resulting in an alteration in transcription and a consequent decrease in enzyme levels. The
investigators did not look for associations between genotype and environmental
exposure. These results suggest that the polymorphism may increase prostate cancer risk

via increasing testosterone levels.

In summary, these studies have shown that polymorphic forms of genes involved in sex
hormone metabolism that can be regulated by PCBs and possibly by dioxin, increase the
risk for two hormone-dependent cancers. As discussed below, endocrine disruption by
environmental contaminants is a major hypothesis for the increasing incidence of

hormone-related cancers as well as increasing male fertility problems.

Some investigators have suggested that male reproductive problems may be linked to
endocrine hormone-disrupting chemicals (86, 105, 235, 295, 296) and OCs have been
shown to disrupt normal endocrine functions (105, 235, 296). The time of exposure to
these chemicals is critical with in utero exposure and exposure before and during puberty
being the most sensitive (297). PCBs can express a mixture of estrogenic, non-estrogenic
and anti-estrogenic effects in in vitro systems (135), depending on the congener involved,
with less chlorinated compounds expressing more estrogenic activity (298). Arochlors
suppressed expression of male-specific CYP genes involved in testosterone metabolism

in rats (299). Dioxins have antiestrogenic effects in rats (136), and can cause partial

83

demasculinization and feminization of sex behavior in male rats exposed in utero (300).
Additionally, a mixture of PCBs and B-estradiol were found to have a synergistic effect
on sex reversal of turtles (30]), suggesting that detrimental interactions may occur

between hormones metabolized by CYP enzymes and environmental contaminants.

2) In genes encoding glutathione S-transferase (GST) enzymes
GSTs are enzymes that catalyze the conjugation of glutathione to electrophilic
xenobiotics in order to inactivate them and facilitate their excretion from the body. They
thus play an important role in detoxifying potentially toxic compounds, including
pesticides and environmental pollutants (reviewed in (281)). However, they have also
been shown to bioactivate some xenobiotics, including the mutagen 1,2-bromoethane
(302). Because of their role in detoxification of genotoxic compounds, most research has

focused on their role in oncogenesis.

Polymorphisms in several GST genes have been identified and characterized. A
polymorphism in GST Mu (GSTMI-I) was identified and found to result in a null
genotype due to deletion of the gene (303). The mutation is most frequently (304),
although not universally (305), associated with an increased risk of lung cancer. It was
also found to slightly increase the risk of postmenopausal breast cancer in the youngest
women (age less than 58 years old) (OR, 2.44 CI, 0.89-6.64) (286). The frequency of the
GSTMl-l genotypes in white Americans is 52% (306). A second GST gene, GST theta
(GS 77’] -I ), which is involved in detoxification of low—molecular weight halogenated

compounds, including several pesticides, was also found to have a null genotype present

84

in 38% of white Americans (307). Like the GSTM1 null mutation, the GSTTl null
genotype may be associated with different types of cancer due to a reduced capacity to
inactivate toxic compounds. Two recently identified polymorphisms in GSTPi (GSTPI-
I) have been identified and one, the GS TPI b allele, is reported to have decreased activity
for compounds such as benzo[a]pyrene diol epoxide and acrolein, carcinogens found in
cigarette smoke (reviewed in (308)). In a study of 155 healthy volunteers from the
Edinburgh area, a total of 6.5% of the individuals were homozygous for the low activity
allele (309). In the same study, the authors also reported on a survey of different types of
cancer that found that subjects with testicular cancer had a prevalence of 18.7% for this
allele vs. 6.5 % for random controls (OR 3.3; CI (1.5-7.7), while a significant decrease in

the GSTPla allele was observed in prostate cancer cases 27.8% vs. 51% for controls

(OR 0.4: C10.02-3.3).

Two studies have investigated the association of polymorphisms in GST genes and
reproductive health. The first study examined the relationship between the GST Mu null
genotype, spermatogenic disorders and alcohol consumption (310). The study involved
271 autopsies of moderate and heavy drinkers, as determined by interviews with
relatives. Spermatogenetic abnormalities were determined by cytological analysis of
biopsied testicular tissue and the GST genotype from cardiac tissue. Heavy drinkers with
the null genotype were less likely to have cytological abnormalities (morphology of the
somniferous epithelium, presence and number of all developmental stages of
spermatozoa) in testicular sections than moderate drinkers. The ORs for disordered

testicular cytology, partial spermatogenic arrest or complete spermatogenic arrest in

85

heavy drinkers with the wild type genotype compared to the variant genotype were 2.0,
(95% CI, 1.0-4.0), 2.0 (95% CI, 0.9-4.2), and 2.0 (CI,0.9-4.1.) These results suggest that
the null genotype was protective against alcohol-induced testicular damage, a surprising
finding since GSTs are usually associated with detoxification pathways. There are several
problems with the study including small study size (50 moderate drinkers and 212 heavy
drinkers), lack of a non-drinking reference group and the inclusion of men with co-
morbidities, such as “other diseases”, drug overdoses and other health problems
associated with heavy alcohol consumption. Additionally the effects of polymorphisms in
CYP 2E1, a gene strongly induced by alcohol, involved in activation of procarcinogens
and consequently associated with ethanol-induced hepatotoxicity (reviewed in (31 l) )
were not investigated. Thus it is difficult to determine the role of the GSTM1

polymorphism in the testicular damage they report.

A second study examined the relationship between polymorphisms in three phase B
detoxification genes, N-acetyltransferase 2 (NAT2), GSTMl and GST’I’I, and
endometriosis (312). Endometriosis is a multifactorial disease with significantly elevated
frequency in industrial areas, possible genetic components and possible associations with
environmental contaminants, including OCs and dioxin (reviewed in (313)). This study
found significant increases in the proportions of patients with minimal and mild
endometriosis and the GSTM1 null genotype compared to controls (75.0% and 79.0%
versus 45.8%, p<0.0001), suggesting that women with the null genotype are at increased
risk for endometriosis. A non- significant increase in the proportion of GSTTl null

genotypes among the patients was also found. Analogous to the modifying role of

86

CYPlAlin the development of postmenopausal breast cancer in PCB exposed women
(282), it seems plausible that a reduced capacity to detoxify reproductive toxicants in
subjects with null mutations in GST genes could play a role in male as well as female

reproductive problems.

In summary, the findings of these reports are suggestive of possible roles of genetic
polymorphisms in genes involved in sex hormone and toxicant metabolism and
reproductive problems in males and females. Clearly more research is needed to better

define these associations.

b. Role of direct toxicity and reactive oxygen species (ROS)
Several researchers have demonstrated a significant relationship between defective sperm
function and oxidative stress (314-318), arising primarily from excessive exposure to
ROS. ROS are produced by leucocytes found in semen, usually as a consequence of
inflammation, or by the sperm themselves during capacitation (319). Evidence for
production of ROS by capacitating spermatozoa was obtained in in vitro experiments
using chemiluminescence probes to detect the superoxide anion. Using a Cypridina
luciferin analog, MCLA, as probe, significant SOD-inhibitable chemiluminescence was
associated with the capacitation of spermatozoa incubated in media supplemented with
fetal cord serum ultrafiltrate. Chemiluminescence was 1270 mV/lO s (with 8 x 106
cells/ml), and corresponded to levels of sperm hyperactivation (12 %) and capacitation

(17%) that were significantly different from those of control spermatozoa (4.9 % and 6

87

%, respectively). The level of capacitation-associated chemiluminescence was directly

related to sperm concentration up to 30 x 106 cells/m1.

Production of ROS by sperm is associated with serious defects in the semen profile,
including severe oligospermia (less than 1 x106 sperm/ml) (320). Support for the role of
ROS produced by the sperm themselves in infertility comes from a study on the semen
quality of oligospermic men. Cells isolated from the ejaculates of a high proportion of
patients exhibiting oligozoosperrnia are characterized by generation rates of reactive
oxygen species that considerably exceed those obtained for the normal fertile population
(Aiken 1992). Semen samples from a cohort of oligozoospermic patients and a group of
fertile controls were fractionated to generate three cell populations of differing density.
For each fraction, both the steady-state and the induced (using a phorbol ester)
chemiluminescent signals were significantly (p< 0.001) greater for the oligozoospermic
samples than for the fertile controls. In the fertile donors, leucocytes comprised the major
source of reactive oxygen species, especially in the low-density fractions; in
oligozoospermic patients, however, spermatozoa were identified as a second major
source of reactive oxygen species. An intense phorbol-ester-induced chemiluminescent
signal generated by purified oligozoospermic spermatozoa, free of leucocyte
contamination, was 167 times greater than the median signal generated by the
corresponding fraction from the fertile controls (p< 0.001). These results emphasize the
importance of spermatozoa as a major source of reactive oxygen species in

01 i gozoospennia.

88

Ortho-substituted but not meta- or para- substituted (i.e. non-coplanar) PCBs have been
shown to activate the oxygen burst in neutophils (321). A PCB congener (3,3’,4,4’—

. tetrachlorobiphenyl) that binds the AhR had no effect on in vitro measures of neutrophil
activation and ROS production while a congener that does not bind (2,2’,4,4’ —
tetrachlorobiphenyl), does (322). Thus the well documented presence of PCBs in seminal

fluid (see above) may affect ROS production by leucocytes present in semen.

Lower chlorinated PCB congeners (1-3 chlorines), although shorter lived than the higher
chlorinated congeners, have been found to persist in environmental samples, biota and in
humans tissues (127) (323, 324), most likely as a result of the dechlorination of the
higher chlorinated congeners. In vitro experiments have shown that the lower chlorinated
congeners are hydroxlated in reactions catalyzed by CYPIA and 2B (127) (325) (326)
and can then be metabolically activated to electrophilic semiquinones and quinones
(McLean et al 1996a). These metabolites can then form PCB- DNA adducts (327) with
the production of superoxide radicals (291). These mechanisms are thought to provide a
basis for the observation that PCB mixtures are complete carcinogens in rodent models
(reviewed in (269). Additionally, production of ROS, including superoxide radicals,
have been shown to damage sperm by inducing lipid peroxidation in the sperm plasma
membrane leading to loss of motility (328), chromatin cross-linking (329), DNA strand

breaks (330) and DNA base oxidation (331).

5. Effects of mercury on male reproduction

89

Studies on experimental animals have shown that administration of methyl mercury to
rodents alters spermatogenesis (332) and results in accumulation of mercury in Sertoli
cells and interstitial tissues of the testes (333). Mercury can induce lipid peroxidation
and ROS generation in rat tissues (334), and liver extracts from rats administered
mercury have reduced CYP2E1 activity (335), possibly as a result of ROS generated

during mercury intoxication.

Several epidemiological studies have found associations of mercury levels in hair or
blood and impaired male fertility. A study on the relationship between Hong Kong male
subfertility and fish consumption found that subfertile males had approximately 40%

more mercury in their hair than did fertile males (336) after adjusting for age.

C. PRELIMINARY STUDIES

1. Previous results from the Fisheaters Family Health Project (FF HP)
The FFHP is an ongoing study funded since 1992 by the CDC that investigates the
association, in licensed Michigan anglers, between exposure to PCBs via Great Lakes
sport-caught fish consumption and male and female reproductive outcomes.

a. Association between fish consumption and PCB levels

The number of sport-caught fish meals consumed in the previous 12 months is a
statistically significant predictor of serum PCB levels for male participants of the FFHP.
A linear model with log10 transformed total PCBs modeled on the dichotomized number
of fish meals in past 12 months (0=less than 12 meals, 1=greater than or equal to 12

meals) was significant (Prob > F = 0.04) and explained 5% of the total variation in serum

90

PCB concentration. A second linear model including age was also significant (Prob > F
=0.02) and explained 28% of the total variation in serum PCB concentration. Adding age
adds so much to the explanation of variance partially because it is confounded with long
term past fish consumption exposures. In the proposed study, we will use the same

categories of fish consumption exposure.

b. PCB congener analysis
PCB congener analyses of 142 male and female participants in the FFHP indicate that of
the 209 possible congeners, 36 were found in the serum of at least one FFHP participant.
Congeners 138/163 (two congeners which co-eluted), 153, 179/190, 180, 187, 194 were
present in at least 30% of the participants, while congeners 138/163 153, 180, and 194
were present in at least 60% of the subjects. These congeners are typical of the congener
profile associated with Great Lakes fish consumption and are consistent with the other
Great Lakes fisheater studies (145, 337). Most of the detected congeners have TEF
values between 10'8 to 10'lo and are not likely to exhibit dioxin-like toxicity (129, 136,
338). Three congeners, 105, 118 and 156, which were detected in 6, 27 and 13%
respectively, of the samples, have TEF values between 10'3 and 10", which is within the

range at which dioxin-like effects are likely to occur (339, 340).

c. Association between PCB levels and reproductive hormone levels
Preliminary analysis of the FFHP data indicates an inverse association between serum
total PCB levels and LH (luteinizing hormone) levels (r=—0.27, p<0.05) in males. Total

PCB values are calculated as the sum of the detected values without substitution for

91

missing values. The levels were not adjusted for lipids. The sample number is small

(N=62), so caution must be used in interpretation of these results.

2. Pilot study: Michigan Fisheaters Study
As part of the FFHP, we tested the feasibility of conducting an exposure-control study on
male fertility problems and exposure to PCBs via Great Lakes sport-caught fish
consumption. We did this in collaboration with Dr. Michael Stahler, the Scientific
Director of the IV F Laboratory, Beaumont Center for Reproductive Endocrinology in
Royal Oak, MI. We collaborated with CDC investigators at the Wisconsin Division of
Public Health, Bureau of Environmental Health (Claire Falk, Henry Anderson, Marty
Kanarek) who are conducting a similar pilot study, in the development of the
questionnaire. We developed the specific protocols and instruments, including a
questionnaire modified from the main male questionnaire from the FFHP, for the study
and determined the most efficient way to recruit participants. We found that having a
receptionist briefly mention the study to clinic clients and then call a staff physician to
talk to the potential participant about the study did not work well. The receptionist was
often uncomfortable or too busy to present the study to the clients, and the physician,
often seeing other clients or supervising laboratory work, was not readily accessible.
Similar results were found by our Wisconsin colleagues and by Dr. Hauser (personal
communications). Due to lack of funds we were not able to hire our own recruiter, as Dr.
Hauser did. This pilot study nevertheless provides the framework for the proposed study,

as described below.

92

3. Research Team

Julia Wirth, PhD MS Principal Investigator. Dr. Wirth was the Project Director
of the FFHP from July 1, 1997 to September 30, 2000. She is currently the Scientific
Coordinator for a large international study on breast cancer in women of Polish ancestry
She has a PhD in microbiology and substantial experience, grant support and publications
in molecular biology, virology, immunology and parasitology. She was a graduate
student in the Department of Epidemiology at MSU from 1996 to 2001, finishing in her
masters degree in May, 2001.

Nigel Paneth, MD, MPH Co-Investigator Dr. Paneth is Chairman of the
Department of Epidemiology, and was the Principal Investigator of the FFHP from July
1, 1997 to October, 1999. He is a pediatrician and perinatal epidemiologist and has been
involved with this project since its inception. He was the recipient of a Great Lakes
Foundation planning grant to develop collaborative research on reproductive effects of
Great Lakes contaminants. Dr. Paneth has extensive research experience in reproductive
and perinatal epidemiology, including the epidemiology of congenital malformations and
perinatal illnesses and the effects of adverse prenatal and perinatal exposures on child
development. Due to his experience in large scale epidemiolgic field studies and analysis
of large data sets, he will be directly involved in data analysis and interpretation of the
results.

Andrew Mullard, BS Project Manager. Mr. Mullard was associated with the
FFHP from 1996 until January, 2001. For the last two years he was a graduate assistant

and was instrumental in developing the questionnaires for the main study and the male

93

reproductive health helped develop the recruiting and tracking protocols and
questionnaire databases. He is currently employed by the Wisconsin Division of Public
Health, Bureau of Environmental Health as Project Director for their ATSDR-funded
study on fish consumption and reproductive health outcomes. Mr. Mullard will be
responsible for the daily conduct of the project, coordinating the efforts of team
members, aiding in manuscript and grant proposal preparation, supervision of student
employees, and organizing and chairing staff meetings.

Jenny Wang, MS Data Analyst. Ms Wang was previously associated with the
FFFP as data analyst and is familiar with the databases and statistical analyses used in
that study. Ms Wang will design and set up the databases for the participant questionnaire
information, hormone analysis, semen analysis and genetic analysis results. She will then
design systems for integrating information from the databases to produce forms and
reports. She will clean the data and perform the statistical analysis.

Michael Diamond, MD Consultant. Dr. Diamond is the Kamran S. Moghissi
Professor of Obstetrics and Gynecology and the Director of Reproductive Endocrinology
and Infertility at the Detroit Medical Center, Wayne State University, Detroit, MI. He
has extensive experience and substantial publications in many aspects of fertility. He will
serve as advisor on the collection and analysis of data from the semen analyses. He will
help oversee the analysis of the semen samples collected from participants at his clinic
and aid in the analysis and interpretation of the results. He will oversee the conduct of
semen analysis for the specific function tests for the PCB subsample study. He will also

lend his expertise on issues regarding male infertility and assessment of outcomes.

94

Harold Humphrey, PhD, Consultant. Dr. Humphrey is a public health scientist
recently retired from the Michigan Department of Community Health but retaining strong
ties with them and active research interests on the effects of PCB exposure and Great
Lakes fish consumption. He has over 25 years of experience in developing and directing
studies examining the human health implications of exposure to environmental chemical
contaminants. He directed the federally funded projects, which established the Great
Lakes fisheater cohort database. He has been a consultant for the FFHP for the last three
years and will provide advise on measuring OC and heavy metal exposures for this
proposal.

Vasantha Padmanabhan, PhD, Consultant. Dr. Padmanabhan is the Senior
Research Scientist and Director of Pediatric Endocrine Research in the Department of
Pediatrics, University of Michigan, and a Senior Research Scientist in the Reproductive
Sciences Program, also at the University of Michigan. She has extensive experience and
publications in the areas of the physiology of reproductive hormone regulation and in the
analysis of human of reproductive endocrine hormones, including inhibin B. Her
laboratory is one of the few that has the capability of measuring specific inhibin subunits
using monoclonal antibody-based assays (ELIZA). She will advise on issues and
interpretation of results of the inhibin B analyses and its relationship to the results of the

other hormones and the semen analyses.

D. RESEARCH DESIGN AND METHODS

1. Overall design

95

We propose to conduct a cohort study in which we will recruit participants based on
exposure to consumption of Great Lakes sport-caught fish. The source population will be
male partners of couples attending infertility clinics in 3 areas of Michigan. For the study
population, we will recruit men based on exposure to consumption to sport-caught Great
Lakes fish. Men who consumed no meals of sport-caught Great Lakes fish in the last year
will be considered unexposed, men who consume 1-11 meals will be considered
moderately exposed and men who consume 12 or more in the last year will be considered
most exposed. We will thus stratify fish consumption into three levels, insuring that we
will have both a highly exposed (12 or more meals), moderately exposed (between 1 and
1 1 meals) and an unexposed (no meals) groups. This strategy, providing a wide range of
exposures, will allow to detect any differences between the fish consumption groups if
they exist. Since we will recruit men from infertility clinics, we also ensure that we will
have the outcome of interest, semen parameters, which include sperm density, sperm
motility and sperm morphology. To investigate possible gene-environment interactions,
we will examine DNA isolated from whole blood samples for polymorphisms in genes
that are involved in contaminant and steroid metabolism. Since the strongest evidence for
an effect of a specific QC on human semen quality comes from the reported adverse
effects of PCBs (217, 218), we will further investigate this relationship with the proposed
cohort. To do this, we will conduct a substudy of men with serum PCB levels in the
upper and lower quartiles of the PCB distribution after we have collected information on
individual serum PCB levels from all the participants. From these men, we will request a

semen sample, which we will use to perform more specific and sensitive tests of sperm

96

function, including the hypoosomotic swelling test, the acrosin release assay and the

zona-binding assay.

a. Study areas
This proposal will focus on three areas of Michigan, each of which abuts on one of
Michigan’s Great Lakes: Allegan, Muskegon and Ottawa counties (Lake Michigan),
Midland, Saginaw and Bay counties (Lake Huron) and St. Claire, Macomb, Wayne and
Monroe counties (Lake Erie)(see the Appendix for a map of the locations). These
counties cover the Great Lakes “areas of concern”, so designated because their waters
contain high levels of persistent and toxic pollutants ( 176, 341). These areas have distinct
as well as overlapping contaminant profiles reflective of the particular industries along
the lakeshore and tributaries and thus provide a spectrum of Great Lakes contaminants.
Since these areas have been the focus of recruitment for the FFHP, we have questionnaire
data, including information on fishing habits and fish consumption, male reproductive
hormone data and PCB levels on 73 men from these areas, which has aided US in
designing the proposed study. The following clinics representing theses areas have agreed
to participate in our proposed study (see the Appendix for their letters of collaboration).
1) Beaumont Center for Reproductive Endocrinology, In vitro Fertilization
Laboratories, Royal Oak, MI. Scientific Director: Michael Stahler;
2) Division of Reproductive Endocrinology and Infertility, the Detroit Medical
Center: Director: Michael Diamond, MD
3) Michigan Reproductive and I VF Center, P. C., Grand Rapids, MI. Louise

Plante, PhD, Director ART Laboratory.

97

4) To be determined, Midland, Saginaw or Bay county

b. Recruitment
Men who come into the clinics for a routine appointment will be approached by a study
recruiter. Those men indicating an interest in participating in the study will be given a
short questionnaire to assess their eligibility and to determine their level of Great Lakes
sport-caught fish in the last year: none, 1-11 meals or 12 or more. The eligibility criteria
include age 18-50 years, not taking hormonal therapy, does not have
diabetes, thyroid or adrenal disorder, does not have genetic disorder related to
fertility, does not have testicular cancer and has not had bilateral orchiectomy. Eligible
me will be given a consent form to sign in which they agree to complete the
questionnaire, donate approximately 20 ml of blood to be used to measure reproductive
hormones and genetic polymorphisms in genes involved in hormone and contaminant
metabolism and give us permission to read their semen analysis report.
We will first recruit an eligible man who has eaten 12 or more meals a year, since, based
on our previous experience, men in this category constitute the lowest percentage of the
general population and mostly likely of the clinic clientele. Once we recruit the highly
exposed man, an eligible moderately exposed man and an unexposed man will be
recruited. Highly exposed men will be matched to moderately and unexposed men on

clinic, age (within 5 years) and time of entry into the study (within 2 months).

This recruitment strategy has two important advantages compared to other studies on fish

consumption and reproductive outcomes. First, by recruiting men from an infertility

98

clinic we will have the semen analysis results, which are sensitive indicators of male
reproductive health (248), and, based on our experience in the FFHP, are often hard to
obtain. We will also have a range of outcomes since the fertility problem has been
reported to be female in a third to a half of the couples (2) (3). Secondly, by recruiting
based on exposure, we ensure that we have some highly exposed men, a population that
is often missing in fisheater studies and may partly explain their inability to find an

association with adverse reproductive outcomes.

We used the following information to estimate that we will be able to recruit about 126
exposed men a year.

Table 1. Estimated recruitment of most exposed men by clinic

 

 

 

 

 

 

Clinic No. No. No. Exposed No. Eating
Samples/year Assuming Assuming >12 fish
80% 20% meals/year
Participation Anglers Assuming
32%
Diamond 420 336 67 22
(Detroit)
Plante l 530 1 224 245 78
(Grand
Rapids)
Stahler 500 400 80 26
(Royal Oak)
TOTALS 2380 1960 392 126

 

 

 

 

 

 

The participation rate is based on the study by Hauser (218), which had rates between 70-
90%, depending on the type of individual doing the recruiting (investigator, nurse) and
time of year the recruiting took place (R. Hauser, personal communication). We will use
a nurse researcher and begin recruiting in the late summer, early fall, 2002, which, based

on Dr. Hauser’s experience, should give us a rate of about 70-80%. Recruiting will be

99

year round in part to randomize any seasonal variations in sperm count that may occur
(18, 67). Based on FFHP cohort A data, we assume an average of 20% of men in
Michigan will obtain yearly fishing licenses. Most anglers in our experience consume
between 5 and 20 fish meals a year. Based on the FFHP cohort A data, 32% of the
anglers consume 12 or more sport-caught fish meals in the last 12 months. We thus
should be able to recruit about 120 exposed and 120 unexposed participants a year for a
total of 376 exposed and unexposed pairs. These calculations were made without
including a possible clinic in the Lake Huron area, which, if included, would increase the

number of exposed subjects recruited by approximately 50 per year.

c. Substudy
Thirty men in the upper and 30 men in the lower quartiles of the PCB distribution from
all three clinics will be selected and asked to donate a semen sample on which to perform
additional sperm function tests. This study will allow us to assess the relationship
between PCB levels and more specific and sensitive measures of sperm function. The
additional sperm function tests will be the: hypoosmotic swelling test, the acrosin assay,
and the zona binding assay. The additional tests have been shown to correlate positively
and significantly with the outcome of in vitro fertilization and sperm fertilizing potential

(342, 343).

d. Questionnaires: development and modification
The questionnaire from the Pilot Study will be used. Prior to administering it to study

participants, it will be field tested for clarity, ease of use and ease and accuracy of data

100

entry on 20 men attending the clinics. The men will also be asked to give their comments
either in writing at the end of the questionnaire or to the study recruiter. Any consistent
comments or recommendations will be used to modify the questionnaire prior to starting

the study.

To assess non-participant bias, we will administer a short questionnaire on fish eating
habits and reproductive health to men who decline to participate. During the three years
of recruiting, approximately 150 randomly selected weeks will be selected during which
all men declining participation will be asked to fill out this questionnaire at the clinics.

We estimate that we will have about 350 completed questionnaires.

e. Training of study recruiters
It has been our experience with the Pilot Study, the experience of our collaborators in
Wisconsin and Dr. Hauser’s experience that having an enthusiastic, trained study
representative dedicated to recruiting participants at the clinic is absolutely critical to
enrolling sufficient participants and thus to the study’s success. We will thus hire a part
time nurse at each clinic whose major responsibility will be to recruit participants. This
person will be present in the clinic from approximately 7 am to 1 pm, which is when the
majority of men come into the clinic (Michael Diamond, personal communication). This
person will be trained by us at MSU regarding the rational of the study, its importance,
how it will be conducted, the requirements and benefits of participation and the purpose
and meaning of the tests and questionnaire. We will stress that the recruiters’ enthusiasm

for the study is critical. This person will also draw the blood sample, oversee data

101

collection, tracking and reimbursement, and communicate weekly with the Project

Manager at MSU.

2. Data Collection
a. Measurement of consumption of Great Lakes fish

Information on consumption of Great Lakes sport caught fish will be obtained from the
questionnaire, which was modified from the FFHP male questionnaire. Men who fish are
asked to identify the water body in which the fish were caught, the fish species caught,
how the fish were prepared, and the amount consumed. This information is critical in
refining the effects of fish consumption on human health outcomes. Both the species and
location of the fish consumed have been shown to be important predictors of total PCB
burdens. We anticipate using this information in models to estimate PCB exposure.
Additional questions about general health, including recent infections with fever,
reproductive health, occupation, medications, smoking habits, and alcohol and caffeine
consumption are also asked. We will also ask about the frequency and duration of hot
tub and Jacuzzi use. We have prepared an identical “core” section of our questionnaire to
be administered both by ourselves and by our Wisconsin colleagues, that includes
questions about fish consumption, medical conditions, risk factors, lifestyle habits,

occupation and environmental exposures and other demographic variables.

b. Other risk factors for infertility

We will exclude men with congenital disorders such as cryptorchidism, even after

repaired by orchipexy, since it is associated with infertility and impaired testicular

102

function (344—346). If men with specific genetic disorders that affect sperm function,
such as Kallman’s syndrome, are identified in the questionnaire, we will exclude them.
Unless varciocele or anti-sperm antibodies contribute to the final diagnosis, we will not

exclude men with these conditions.

Clinically apparent infections of the testes, epididymis, or the prostate can lead to
reduced sperm count or motility (reviewed in (19). The physical examination the men
undergo as part of the fertility work-up will identify men with these conditions. To
identify past infections or recent subclinical infections that might affect spermatogenesis,
we ask in our questionnaire if the men have ever had specific diseases and conditions that
are associated with impaired sperm function. We also ask if the men have had any viral
or bacterial infections or flu that caused a fever in the last 3 months prior to semen
collection. Thus, we will identify men with these possible risk factors but will not

exclude them.

Smoking has clearly been associated with a variety of sperm abnormalities in several
studies (342, 347-349). In our questionnaire we ask about smoking, both the type of
tobacco used and the quantity, and will control for it in the analysis. Chronic alcohol
abuse can lead to testicular atrophy (350), while little consistent effect on sperm function
has been found for caffeine consumption (351). We ask about consumption of both in the

questionnaire and will control for them if necessary.

c. Laboratory tests

103

1) Reproductive hormones
To obtain as accurate and as sensitive a measure of spermatogenesis as possible, we will
assay the serum samples for inhibin B and estradiol, as well as for FSH, LH and
testosterone (TE). Serum inhibin B level correlated with sperm concentration in several
studies (352-354) and is considered to be the best available endocrine marker of
spermatogenesis in subfertile men (355). The two-site ELISA assay for inhibin B (355,
356) uses a monoclonal antibody able to specifically detect and distinguish between the
two bioactive forms of inhibin, A and B. The assay will be performed in the laboratory of
Dr. Vasantha Padmanabhan at the University of Michigan who has extensive experience

using it in large scale studies (please see the Appendix for her letter of collaboration).

Spermatogenesis is controlled by the hypothalamic-pituitary-testicular axis (reviewed in
de Kretser (357). Gonadotropin-releasing hormone (GnR) stimulates the pituitary gland
to release FSH and LH. FSH in the male, in combination with TE, is required for the
initiation and maintenance of spermatogenesis. The primary target of FSH is the Sertoli
cell, which provides the nutrients needed by the stem cells in the somniferous tubules to
differentiate into mature motile sperm. Sertoli cells also secret inhibin B, which inhibits
FSH secretion by the anterior pituitary cells. LH, also produced by the anterior pituitary,
regulates the function of Leydig cells in the interstitial tissues of the testis. Leydig cells
produce TE and may also utilize it. TE inhibits LH secretion by the anterior pituitary and
may also negatively hormone secretion by the hypothalamus. Estradiol in adult men is
required for synthesis of sex hormone binding globulin, which controls the level of

available TE in the circulation, and helps regulate gonadotropin secretion (358). FSH,

104

LH, TE and estradiol will be measured from serum samples, using routine laboratory

methods.

Secretion of FSH, LH, and TE is subject to biological rhythms, but there is no consensus
as to their effects on serum hormone levels. TE levels have been reported to decrease
during the day (359), although the variation diminishes with age (360). The secretory
patterns of FSH and LH are pulsatile, but no clear pattern has been found (361, 362).
Another study found that differences in FSH and inhibin B levels could not be explained
by sampling time (353). The participating clinics report that most men come in between
early morning and early afternoon, which should help reduce any possible variation due
to sampling time. Additionally, we will record the time of day the serum samples are
drawn and assess its effect on hormone levels in the analysis. To ensure comparability
between the clinics, random serum samples from participants will be aliquoted and
assayed in each lab. Kappa statistics will be calculated to evaluate interobserver

agreement.

2) Semen analysis
Although the clinical value of the traditional semen parameters in the diagnosis of
male fertility has been debated recently (363), several studies have found that the
measured parameters are predictive of pregnancy outcome (364) (363) as well as for the
likelihood of successful in vitro fertilization (reviewed in (365)). Routine semen analyses
measures sperm number, sperm motility, sperm morphology as well as semen

characteristics, such as liquification, volume, viscosity, pH and number of white blood

105

cells. Semen analysis is routinely performed in all the participating clinic laboratories as
part of the fertility work-up. All of the clinics follow the procedures recommended by
the World Health Organization (15) and have rigorous internal quality control measures
(periodic analyses of the same sample by the technicians and multiple analyses of the

same sample by the same technician).

To ensure comparability of the laboratory analyses, prior to recruitment test samples from
volunteers will be analyzed at each laboratory. A portion of the semen will be washed
and resuspended in buffered 1% formalin (1 1). Several dilutions will be made and
aliquoted into 3 sets of cryovials. The vials will be transported by courier to each of the
three labs for analysis of sperm density. For analysis of morphology, slides will be
prepared from the samples and sent to the laboratories. For motility analysis, the semen
samples will be aliquoted in to 3 cryovials and frozen at —1800C. The vials will be
transported by the same courier as for the semen samples for sperm counting to the three
laboratories (11). Interobserver agreement between the different laboratories will be
assessed using the Kappa statistic. Additionally, the Detroit Medical Center has the
capability of videotaping semen samples. The actual semen material can then be viewed
at the other laboratories and sperm count and motility can be analyzed. We will also test
the feasibility of this procedure for external quality control. After the study has begun,
random samples will be collected at each clinic and then analyzed by all 3 clinics.
Interobserver agreement will be assessed using the Kappa statistic. All laboratory

personnel in the clinics will be blinded as to the identity and level of fish consumption of

106

the participants. We will also have access to information on the final diagnosis and if the

problem is male or female when it becomes available.

3) Additional sperm function tests
Ideally for the substudy, sperm function tests would be chosen to test each of the
critical reactions a sperm must undergo to successfully fertilize an ovum. Twelve such
steps have been described (343). While a plethora of such tests are available (reviewed in
(343) (366), many of them are not accurate or reproducible, or are too expensive for large
scale studies such as this one (365). For this study, three tests were chosen to reflect
specific critical reactions that correlate with in vitro fertilization potential and sperm
fertilizing capacity (342) (365) that can be measured accurately and reproducibly. The
additional tests will be performed under the supervision of Dr. Diamond at the Detroit

Medical Center.

The hypoosomotic swelling test is a test of sperm survival and membrane integrity and
will be assessed by a dye exclusion test. This test has been reported to be useful in
assessing male fertility (367). Before sperm can penetrate the ovum, it must undergo
several physiological changes in the female genital tract. The first is capacitation, a series
of biochemical and functional changes involving removal and redistribution the
membrane constituents of the sperm surface. Following this, the acrosome, a membrane
enclosed structure filled with enzymes at the sperm head, reacts resulting in the
dissolution of the outer acrosome membrane and the release of the acrosomal contents,

one of which is the lytic enzyme acrosin. An additional consequence of capacitation is a

107

Change in the sperm tail beat pattern to one of hyperactivity. After capacitation and the
acrosome reaction, the sperm are able to attach to receptors on the zona pellucida of the
ovum and penetrate it. We will use two tests to evaluate steps in these processes. To
evaluate the acrosome reaction, we will use a test to detect released acrosin (368). To
evaluate sperm binding to the zona pellucida, we will use the zona-binding assay,
possibly with recombinant zona proteins (365).

4) Genetic analysis
We are collecting a whole blood sample from which to isolate genomic DNA. In
collaboration with Dr. Michael Shi at Parke-Davis Pharmaceutical Research in Ann
Arbor, MI, we will look for polymorphisms in 11 genes associated with contaminant,
steroid and/or drug metabolism (see the Appendix for his letter of collaboration). The
genes with polymorphisms that are relevant to this proposal (please see Section C.4.c.)
are CYP1A1, CYP3A4, GST Mu and GST Theta. In addition, Parke-Davis is interested
in determining the gene frequency in our study population of polymorphisms in
CYP2D6 (debrisoquine hydroxylase), NQOl (NAD(P)H (quinone oxidoreductase),
SULTI (STP1)(phenol sulfotransfease), CYP2C9 (tolbutamide methylhydroxylation),
and CYP2C19 (S-mephenytoin 4’-hydroxylation), N-acetyltransferase (NAT2) because
these genes are also involved in metabolism of drugs. DNA will be isolated using
standard protocols and the sequences of interest amplified using polymerase chain
reaction technology with the appropriate primers. A portion of each isolated DNA
sample will be frozen and stored for future analysis.

5) Contaminant analysis

a) PCBs

108

Serum samples will be analyzed for 208 PCB congeners, DDT, DDE and several other
pesticides for which the laboratory routinely screens. The analyses will be performed at
Michigan Department of Community Health (MDCH), Lansing, MI, which has
performed identical analyses for the FFHP, as well as for most other Fisheater studies
conducted in Michigan (145, 208). For PCB analysis, the laboratory uses capillary
column gas chromatography to identify the specific PCB congeners, DDT and DDE. The
detection sensitivity of this method varies according to congener, ranging from 0.03 to
1.30 ppb. Aliquots of the remaining sera will be stored frozen and will be available for
future determinations of other pesticides and industrial contaminants. The MDCH
laboratory adheres to strict internal and external QA/QC practices and has over twenty
years of experience in developing and conducting chemical analyses on human
specimens. The analyses will be performed in a comparable manner to those conducted
on previous and contemporary cohorts to provide full comparability to previous results
and permit longitudinal as well as geographic comparisons. We are aware that serum
PCB levels may require correction for serum lipid level(369). Although serum PCB
levels do appear to correlate well with adipose tissue PCB levels (370), enough blood

will also be drawn for serum lipid analysis in order to correct for lipid levels as required.

b) Dioxin
Dioxin analysis will be performed by Xenobiotic Detection Systems, Inc (XDS)(please
see the Appendix for their letter of collaboration). XDS uses a patented chemical
activated luciferase (CALUX) gene expression assay to screen for persistent

bioaccumulating toxins, including dioxins and PCBs. The assay is based on toxicant

109

binding to the aryl hydrocabon receptor (AhR), which has been attached to the firefly
luciferase gene and thus controls its expression. A second step then separates PCBs,
dioxins and dibenzofurans. The assay routinely measures dioxins in low parts per trillion.
A study measuring dioxin concentrations found a positive correlation (r=0.8l) between
this method and the gas chromatography/mass spectrometry (GC/MS) method (371). The
advantages of this system compared to GC/MS are its low cost and high sensitivity.

Samples that are positive for dioxins using this method will be saved for further analysis

using GS/MS, if warranted.

c) Mercury
To detect mercury, whole blood samples will be analyzed using a modification of the
cold vapor atomic adsorption spectrophotometry method. The MDCH lab routinely
analyzes samples for mercury using this method. The laboratory will also perform an
analysis for lead at no extra cost. The results of this analysis will be recorded. and stored
in a database, but not evaluated as part of this study. While lead exposure has been
consistently associated with male reproductive problems, including decreased sperm
concentrations and total sperm counts, poor sperm motility, and abnormal sperm

morphology (372) (373), there are many sources of exposure beside fish consumption.

3. Data analysis and interpretation
a. Data analysis
The aims of this study are both descriptive and comparative. Information on several

exposure variables and a large number of outcome variables, including covariates, will be

110

collected on men attending infertility clinics in four areas of Michigan. Exposure to
consumption of sport-caught Great Lakes fish is the primary exposure. Secondary
exposures are the contaminants found in these fish: PCBs, DDE, dioxin, and mercury.
The primary outcome is sperm density, which is one of the group of outcome variables
obtained from the semen analysis. Other semen variables include semen volume, sperm
motility, and sperm morphology. The second group of outcome variables include the
male reproductive hormones FSH, LH, testosterone and inhibin B. The outcome variables
will first be inspected as continuous variables and then may be categorized as

appropriate.

The descriptive statistics will be used to characterize the cohort and will include
distributions of age, BMI, area of recruitment, fish consumption level, and contaminant
levels, including total and congener-specific levels PCB level, and reproductive hormone

levels. Distribution of the genetic polymorphisms will also be determined.

For the univariate analyses of high and low fish consumers matched on age and
recruitment area, all variables will be categorical. Comparisons of the pairs with the
outcomes will be made using rate ratios derived from 2 x 2 contingency tables. For
analysis of the variable as continuous, t-tests and correlational methods will be used. If
the variable is not normally distributed, which is likely for sperm densities (4) (6) and
PCB values (145), the data will be log-transformed. If the log-transformed values are not
normally distributed, non—parametric tests, such as the Wilcoxon rank sum test and rank

correlations methods will be used. Since PCB levels usually have a wide distribution,

11]

PCB exposure levels may be divided into qu'antiles. While the analytic procedure used by
MDCH can theoretically detect 209 PCB congeners, only 33 peaks, representing 37

con geners (four peaks were co-eluting pairs) were detected in the serum from FFHP
cohort A participants. However, since multiple comparisons can generate significant
relationships by chance alone, multiple-comparison procedures will be used if
appropriate. For statistical analysis, PCB congeners may be grouped into various
categories including: chlorination clusters (374), isoforms, co-planar and ortho-
substituted congeners, and the most prevalence congeners. Logistic regression models
will be used to ascertain the contribution of these con gener groups to any detected

adverse health effects.

Exposure variables found to be significantly associated with an outcome will be
incorporated into a multivariate regression models, logistic for categorical variables and
linear for continuous variables. Included in these models will be various covariates,
including BMI, infections, and occupation. A model using species and location as
predictors of ingested dose will also be tested (375). The final model will include the

significant exposure variables as well as potential confounders.

Interaction between fish consumption and the secondary exposure variables and genetic

polymorphisms will tested using statistical models designed to examine interactions.

b. Sample size/power calculation

112

The sample size was calculated for two independent means for differences in sperm
density. Alpha was set at 0.05, power at 0.80, the true group mean under the null
hypothesis at 100.8 x 106 (million) sperm/ml, the smallest group 2 mean at 80.64, the
maximum group 2 mean at 88.0 million/ml, and the standard deviation was calculated at
72.0 million/ml (48). The sperm count values were taken from a report on semen
analyses on US men who banked sperm prior to vasectomy in the years 1970 to 1994.
The specific values used in the above calculations were from 662 men that donated their

samples at Cryogeneic Laboratories, Inc., Roseville, MN.

Table 2. Sample size requirements

SAMPLE SIZE REQUIREMENTS FOR COMPARING TWO
MEANS where Alpha=.05, Sides=2, STD=72.0, Power=.80,
Ratio N2/N 1:2, Group 1 true mean=100.8(million/ml)

Group 1 Group 2
Mean (million/m1) Mean (million/ml) N1 N2
100.8 80.64 151 302
100.8 81.14 159 318
100.8 81.64 167 334
100.8 82.14 176 352
100.8 82.64 186 372
100.8 83.14 197 394
100.8 83.64 208 416
100.8 84.14 221 442
100.8 84.64 235 470
100.8 85.14 250 500
100.8 85.64 267 534
100.8 86.14 285 570
100.8 86.64 306 612
100.8 87.14 328 656
100.8 87.64 354 708
100.8 88.14 382 764
100.8 88.64 414 828
100.8 89.14 450 900
100.8 89.64 491 982

113

Power is calculated with the same values for the variables as above. With a sample size
of 378 most—exposed men, 378 moderately exposed men and 378 unexposed men
(N2=752) and we have 80% power to detect a difference between means of 11.2 million
sperm/m1.

Table 3. Power estimates

POWER ESTIMATES FOR TESTING HYPOTHESIS WHERE
Alpha=.05, Sides=l, Pooled SD=72.0, N 1:378,
N2=756, Group 1 true mean=100.8(million/ml)

Group 1 Group 2

Mean (million/ml) Mean (million/ml) POWER
100.8 80.64 0.99740
100.8 81.14 0.99636
100.8 81.64 0.99497
100.8 82.14 0.99313
100.8 82.64 0.99071
100.8 83.14 0.98757
100.8 83.64 0.98356
100.8 84.14 0.97849
100.8 84.64 0.97215
100.8 85.14 0.96433
100.8 85.64 0.95479
100.8 86.14 0.94330
100.8 86.64 0.92962
100.8 87.14 0.91354
100.8 87.64 0.89486
100.8 88.14 0.87342
100.8 88.64 0.84911
100.8 89.14 0.82187
100.8 89.64 0.79173

c. Project timetable
Prior to beginning recruiting, we will conduct the external quality control study. When
we are satisfied that minimal differences exist between the laboratories (kappa above
0.75) will begin recruiting subjects. Since the instruments and protocols have already
been developed, excluding the substudy, we will begin recruiting within 3 months of

receiving funding. During that time, we will also seek approval from the participating

114

institutions’ internal review boards (IRBs), select and train the half time personnel
responsible for recruiting at each clinic, set up databases, develop computer linkages for
data entry and pilot test the questionnaire. Based on the calculations above (please see
section D.2.b.), we estimate it will take approximately 3 years to recruit the 800 subjects.
During that time, questionnaire data will be collected, cleaned and entered into the
databases. Contaminant, hormone and genetic analyses will be performed and the results
entered into the appropriate databases. Data analysis will begin when this is complete and

continue through the end of the funding period.

6. Strengths and weaknesses
a. Strengths

This study has several strengths that set it apart from other studies examining the
relationship between environmental contaminants and male reproductive health. First, our
study subjects are men who attend infertility clinics, and thus a higher percentage are
likely to have fertility problems than in an unselected population sample. It is estimated
that approximately one third to one half of couples’ infertility problems will be male (3)
(2). We will have access to their semen analysis results as well as the final diagnosis. We
will also perform sperm function tests in a subsample of men exposed to high and low
levels of PCBs in order to detect more subtle changes in sperm functions as a
consequence of PCB exposure. Second, we are recruiting subjects based on exposure.
This will ensure that we have a range of exposures, including highly exposed subjects,

thus increasing our chances of detecting adverse effects if they exist. Third, we will look

115

for polymorphisms in the subjects’ genes involved in contaminant and sex steroid
metabolism that are known to alter toxicant metabolism and ultimately concentration.
b. Weaknesses

Sampling from an infertility clinic creates a risk for selection bias. Not all couples
with infertility problems seek medical help (376). If factors related to the decision to
seek medical help are also related to our exposure, Great Lakes fish consumption, this
may result in etiologically irrelevant differences in exposure between infertile care-
seeking cases and fertile controls who did not seek medical help. However, this is
unlikely since there is no a priori reason to associate care—seeking behavior and fish
consumption. Selection patterns might differ according to the type of procedures the
couples select. To minimize the potential for selection bias, we will recruit more and less
exposed men from the same group of couples coming in for consultation. Ideally we
would chose couples having their first consultation only, but based upon the experience
of another similar ongoing study by Dr. Hauser (218), this approach would eliminate a
significant number of potential participants. It is also possible that men attending
infertility clinics may represent a more susceptible population, which would potentially
influence the relationship between exposure and outcome. However, studies comparing
sperm counts from infertility clinics and from semen donors have not found great

differences (4).

We recognize that there is significant day to day variability in semen parameters. This
misclassification bias should be nondifferential and reduces power because it reduces our

ability to discriminate between men with normal and abnormal semen parameters, but

116

only biases our results toward the null. Ideally we would collect more than one sample
but due to the large sample size (756 men) this would make the study prohibitively
expense. We have some limited data from the FFHP on men who donated a semen
sample that was found to be abnormal and were subsequently asked to donate a second
one. In 4 of 5 cases, even though there were variations in the values, the abnormal values

remained within the abnormal range.

E. HUMAN SUBJECTS

Participants in the proposed study will be adult (18-50 year old) males attending
infertility clinics in Michigan. All potential participants will receive an informed consent
statement for their consideration and signature before any additional information or
samples are obtained. Participants will be asked to complete a questionnaire on
reproductive history, fish consumption, occupation and other factors potentially
confounding the relationship between contaminant exposure and semen parameters. Men
will be asked to have their blood drawn on one occasion and the risks to participants are
minimal. The phlebotomy will be performed by technicians employed by the clinics and
will be in compliance with Michigan licensing laws. A subsample of men will be asked to
donate a semen sample for this study for which there are minimal risks. All data will be
maintained under confidential conditions as mandated by Michigan State University.
Participants will be tracked using the clinic study identifiers unrelated to personal
information. Personal identifiers will be stored locked and separate from the study data.
All data from the original forms will be coded and entered into a computer file, and the

coded computer file will be used for all data analyses. Participants will receive several

117

incentives to increase participation. First, they will receive the results of their own tests

for contaminants, reproductive hormones, semen quality and genetic analysis. Second,

they will receive a cash incentive of $50 after completion of the phlebotomy and the

questionnaire. All test results will be made available to the study participants’ physicians

upon request.

F. REERENCES

l.

NORA. Fertility and pregnancy abnormalities: NIOSH National Occupational
Research Agenda (NORA), 1999.

Skakkebaek NE, Giwercman A, de Kretser D. Pathogenesis and management of
male infertility [see comments]. Lancet 1994;343:1473-9.

De Kretser DM, Baker HW. Infertility in men: recent advances and continuing
controversies. J Clin Endocrinol Metab 1999;84:3443-50.

Carlsen E, Giwercman A, Keiding N, Skakkebaek NE. Evidence for decreasing
quality of semen during past 50 years [see comments]. ij 1992;305:609-13.

Brake A, Krause W. Decreasing quality of semen. ij 1992;305:1498.

Bromwich P, Cohen J, Stewart 1, Walker A. Decline in sperm counts: an artefact
of changed reference range of "normal"? [see comments]. ij 1994;309:19—22.

MacLeod J, Wang Y. Male fertility potential in terms of semen quality: a review
of the past, a study of the present. Fertil Steril 1979;31:103-16.

Kruger TF, Menkveld R, Stander FS, et al. Sperm morphologic features as a
prognostic factor in in vitro fertilization. Fertil Steril 1986;46:1118-23.

Bonde JP, Giwercman A, Ernst E. Identifying environmental risk to male
reproductive function by occupational sperm studies: logistics and design options.

Occup Environ Med 1996;53:511-9.

Sheriff DS. Analysis of semen in a constantly changing social context of
medicine. Arch Androl 1995;34:125—32.

118

ll.

12.

l3.

14.

15.

l6.

l7.

19.

20.

21.

22.

23.

24.

Neuwinger J, Behre HM, Nieschlag E. External quality control in the andrology
laboratory: an experimental multicenter trial. Fertil Steril 1990;54:308-14.

Matson PL. External quality assessment for semen analysis and sperm antibody
detection: results of a pilot scheme. Hum Reprod 1995;10:620-5.

Mallidis C, Howard EJ, Baker HW. Variation of semen quality in normal men. Int
J Androl 1991;14:99-107.

Schrader SM, Turner TW, Simon SD. Longitudinal study of semen quality of
unexposed workers. Sperm motility characteristics. J Androl 1991;12:126-31.

Organization WHO. WHO laboratory manual for the examination of human
semen and sperm-cervical mucus interaction. Cambridge, MA: World Health
Organization, 1999.

Mees MM, Cuijpers CEJ, Piersma AH. Is sperm quality actually declining? A
literature survey. Bilthoven, the Netherlands: National Institute of Public health
and Environment, 1997.

MacLeod J, Gold R2. The male factor in fertility and infertility. H. Spermatozoon

counts in 1000 men of known fertility and in 1000 cases of infertile marriage. J
Urol 1951;66:436-49.

Tjoa W. Circannual rhythm in human sperm count revealed by serially
independent sampling. Fertil Steril 1982;38:454-9.

Clark R. Male infertility. In Principles and practice of endocrinology and
metablism, second edition, ed: K.L. Becker, J.B.Lippincott Co., Philadelphia;
pp.1102-1110. 1995.

Tielemans E, Heederik D, Burdorf A, Loomis D, Habbema DF. Intraindividual
variability and redundancy of semen parameters. Epidemiology l997;8:99-103.

Rock J, Robinson D. Effect of induced intrascrotal hyperthermia on testicular
function in man. Am J Obstet Gynecol 1965;93:793-801.

Mieusset R, Bujan L. Testicular heating and its possible contributions to male
infertility: a review. IntJ Androl 1995;18:169-84.

Brindley GS. Deep scrotal temperature and the effect on it of clothing, air
temperature, activity, posture and paraplegia. Br J Urol 1982;54:49-55.

Tiemessen CH, Evers JL, Bots RS. Ti ght-fitting underwear and sperm quality.
Lancet 1996;347: 1844-5.

119

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

37.

38.

Sanger WG, Friman PC. Fit of underwear and male spermatogenesis: a pilot
investigation. Reprod Toxicol 1990;4z229—32.

Bujan L, Mieusset R. [Male contraception by hyperthermia]. Contracept Fertil
Sex 1995;23:611-4.

Lynch R, Lewis-Jones DI, Machin DG, Desmond AD. Improved seminal
characteristics in infertile men after a conservative treatment regimen based on
the avoidance of testicular hyperthermia. Fertil Steril 1986;46:476-9.

Gyllenborg J, Skakkebaek NE, Nielsen NC, Keiding N, Giwercman A. Secular
and seasonal changes in semen quality among young Danish men: a statistical

analysis of semen samples from 1927 donor candidates during 1977-1995. Int J
Androl 1999;22:28-36.

Levin RM, Latimore J, Wein A], Van Arsdalen KN. Correlation of sperm count
with frequency of ejaculation. Fertil Steril 1986;45:732-4.

Mortimer D. The male factor in infertility. Part I. semen analysis. Curr Probl
Obstet Gynecol Fertil 1985;8z3-87.

Pellestor F, Girardet A, Andreo B. Effect of long abstinence periods on human
sperm quality. Int J Fertil Menopausal Stud 1994;39:278-82.

Schwartz D, Mayaux MJ, Guihard-Moscato ML, et a1. Study of sperm
morphologic characteristics in a group of 833 fertile men. Andrologia
1984;16:423-8.

Bendvold E, Gottlieb C, Bygdeman M, Eneroth P. Depressed semen quality in
Swedish men from barren couples: a study over three decades. Arch Androl

1991;26:189-94.

Nelson CM, Bunge RG. Semen analysis: evidence for changing parameters of
male fertility potential. Fertil Steril 1974;25:503-7.

Westoff CF. Coital frequency and contraception. Fam Plann Perspect l974;6: 136-
41.

James WH. What stabilizes the sex ratio? Ann Hum Genet 1995;59:243—9.
James WH. Secular trend in reported sperm counts. Andrologia 1980;12:381-8.

Seracchioli R, Porcu E, Flamigni C. The diagnosis of male infertility by semen
quality. Sperm morphology is not the only criterion of male infertility. Hum
Reprod 1995;10:1039-41.

120

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

Organization WHO. WHO laboratory manual for the examination of human
semen and sperm-cervical mucus interaction. Geneva: WHO, 1987.

Chan SY, Wang C, Chan ST, et a1. Predictive value of sperm morphology and
movement characteristics in the outcome of in vitro fertilization of human
oocytes. J In Vitro Fert Embryo Transf 1989;6:142-8.

Olsen GW, Bodner KM, Ramlow J M, Ross CE, Lipshultz LI. Have sperm counts
been reduced 50 percent in 50 years? A statistical model revisited [see
comments]. Fertil Steril 1995;63:887-93.

Lerchl A, Nieschlag E. Decreasing sperm counts? A critical review. Exp Clin
Endocrinol Diabetes 1996;104:301-7.

Wang C, Chan SY, Leung A, et al. Cross-sectional study of semen parameters in a
large group of normal Chinese men. Int J Androl 1985;82257-74.

MacLeod J aHL. Characteristics and variations in semen specimens in 100 normal
young men. J Urol 1945;54:474-82.

Keiding N, Giwercman A, Carlsen E, Skakkebaek NE. Falling sperm quality
[letter; comment]. ij 1994;309: 131.

Fisch H, Goluboff ET. Geographic variations in sperm counts: a potential cause of
bias in studies of semen quality. Fertil Steril 1996;65: 1044-6.

Leto S, Frensilli FJ. Changing parameters of donor semen. Fertil Steril
1981;36:766-70.

Fisch H, Goluboff ET, Olson JH, Feldshuh J, Broder SJ, Barad DH. Semen
analyses in 1,283 men from the United States over a 25-year period: no decline in

quality [see comments]. Fertil Steril 1996;65:1009-14.

Paulsen CA, Berman NG, Wang C. Data from men in greater Seattle area reveals
no downward trend in semen quality: further evidence that deterioration of semen
quality is not geographically uniform [see comments]. Fertil Steril 1996;65:1015-
20.

Saidi JA, Chang DT, Goluboff ET, Bagiella E, Olsen G, Fisch H. Declining
sperm counts in the United States? A critical review. J Urol 1999;161:460—2.

Bostofte E, Serup J, Rebbe H. Relation between spermatozoa motility and
pregnancies obtained during a twenty-year follow-up period spermatozoa motility
and fertility. Andrologia 1983;15:682-6.

121

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

Irvine S, Cawood E, Richardson D, MacDonald E, Aitken J. Evidence of
deteriorating semen quality in the United Kingdom: birth cohort study in 577 men
in Scotland over 1 1 years [see comments]. ij 1996;312:467—71.

Auger J, Kunstmann JM, Czyglik F, Jouannet P. Decline in semen quality among
fertile men in Paris during the past 20 years [see comments]. N Engl J Med
1995;332:281-5.

Bujan L MA, Pontonnier F and Mieusset R. Time series analysis of sperm
concentration in fertile men in Toulouse, France between 1997 and 1992. British
Medical Journal 1996;312:471-2.

Ginsburg J, Hardiman P. Ovulation induction with human menopausal
gonadotropins--a changing scene. Gynecol Endocrinol 1991 ;5:57-78.

Ginsburg J, Okolo S, Prelevic G, Hardiman P. Residence in the London area and
sperm density. Lancet 1994;343:230.

Farrow S. Falling sperm quality: fact or fiction? ij 1994;309:1-2.

Bahadur G, Ling KL, Katz M. Statistical modelling reveals demography and time
are the main contributing factors in global sperm count changes between 1938 and
1996. Hum Reprod 1996;] 1:2635-9.

Swan SH, Elkin EP, Fenster L. Have sperm densities declined? A reanalysis of
global trend data. Environ Health Perspect 1997;105:1228-32.

Swan SH, Elkin EP, Fenster L. The question of declining sperm density revisited:
an analysis of 10] studies published 1934-1996. Environ Health Perspect
2000;108:961-6.

Bostofte E, Serup J, Rebbe H. Has the fertility of Danish men declined through
the years in terms of semen quality? A comparison of semen qualities between
1952 and 1972. Int J Fertil 1983;28:91-5.

Osser S, Liedholm P, Ranstam J. Depressed semen quality: a study over two
decades. Arch Androl l984;12:1 13-6.

Menchini-Fabris F, Rossi P, Palego P, Simi S, Turchi P. Declining sperm counts
in Italy during the past 20 years. Andrologia 1996;28:304.

Bonde JP, Kold Jensen T, Brixen Larsen S, et a1. Year of birth and sperm count in
10 Danish occupational studies. Scand J Work Environ Health 1998;24:407-13.

Fisch H, Ikeguchi EF, Goluboff ET. Worldwide variations in sperm counts.
Urology 1996;48:909-1 1.

122

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

Bendvold E. Semen quality in Norwegian men over a 20-year period. Int J Fertil
1989;34:401-4.

Vierula M, Niemi M, Keiski A, Saaranen M, Saarikoski S, Suominen J. High and
unchanged sperm counts of Finnish men. Int J Androl 1996;19:11-7.

Bostofte E, Serup J, Rebbe H. Interrelations among the characteristics of human
semen, and a new system for classification of male infertility. Fertil Steril
1984;41:95-102.

Bostofte E, Serup J, Rebbe H. Relation between morphologically abnormal
spermatozoa and pregnancies obtained during a twenty-year follow-up period. Int
J Androl l982;5:379-86.

Bostofte E. Prognostic parameters in predicting pregnancy. A twenty-year follow-
up study comprising semen analysis in 765 men of infertile couples evaluated by
the Cox regression model. Acta Obstet Gynecol Scand 1987;66:617-24.

Bonde JP, Ernst E, Jensen TK, et a1. Relation between semen quality and fertility:
a population-based study of 430 first-pregnancy planners [see comments]. Lancet
1998;352:1172-7.

Chilvers C PM, Forrnan D, Fogelman K, Wadsworth MEJ. Apparent doubling of
frequency of undescended testis in England and Wales in 1962-81. Lancet
1984;8398:330-332.

Scorer C. The descent of the testis. Arch Dis Childh 1964;39:605-609.

Calzolari E CM, Roncarati E, et al. Aetiologic factors in hypospadias. J Med
Genet 1986;23:333-337.

Paulozzi LJ, Erickson JD, Jackson RJ. Hypospadias trends in two US surveillance
systems. Pediatrics 1997;100:831-4.

Dolk H. Rise in prevalence of hypospadias. Lancet 1998;351:770.

Ansell PE BV, Bull D et al,. Cryptorchidism: a prospective study of 7500
consecutive male births, 1984-8. Arch Dis Child 1992;67:892-899.

Group J RHCS. Cryptorchidism: a prospectivestudy of 7500 consecutive male
births, 1984-8. Arch Dis Child 1992;67:892-899.

Berkowitz GS, Lapinski RH, Dolgin SE, Gazella JG, Bodian CA, Holzman IR.
Prevalence and natural history of cryptorchidism. Pediatrics 1993;92:44-9.

123

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

Matlai P, Beral V. Trends in congenital malformations of external genitalia.
Lancet 1985;]:108.

Aho M, Koivisto AM, Tammela TL, Auvinen A. Is the incidence of hypospadias
increasing? Analysis of Finnish hospital discharge data 1970-1994. Environ
Health Perspect 2000;108:463-5.

Systems ICFBDM. Annual Reports 1980-1993. Rome: International Center for
Birth Defects, 1982-95.

Group EW. 15 years of surveillance of congenital anomalies in Europe 1980—
1994. Brussels: Scientific Institute of Public Health-Louis Pasteur, 1997.

Canning DA. Hypospadias trends in two US surveillance systems. Rise in
prevalence of hypospadias. J Urol 1999;161:366.

Choi J, Cooper KL, Hensle TW, Fisch H. Incidence and surgical repair rates of
hypospadias in New York state. Urology 2001;57:151-3.

Toppari J, Larsen JC, Christiansen P, et a]. Male reproductive health and
environmental xenoestrogens [see comments]. Environ Health Perspect 1996;104
Suppl 4:741-803.

Nethersell AB, Drake LK, Sikora K. The increasing incidence of testicular cancer
in East Anglia. Br J Cancer 1984;50:377-80.

Pike MC, Chilver CED, Bolerow LG. Classification of testicular cancer in
incidence and mortality statistics. Br J Cancer 1987;56:83-85.

Boyle P, Kaye SB, Robertson AG. Changes in testicular cancer in Scotland. Eur J
Cancer Clin Oncol 1987;23:827-30.

Hakulinen T, Andersen A, Malker B, Pukkala E, Schou G, Tulinius H. Trends in
cancer incidence in the Nordic countries. A collaborative study of the five Nordic
Cancer Registries. Acta Pathol Microbiol Irnmunol Scand Suppl 1986;288:1-151.

Adami HO, Bergstrom R, Mohner M, et al. Testicular cancer in nine northern
European countries. Int J Cancer 1994;59:33-8.

Stone J, Sandeman TF, Ironside P, Cruickshank DG and Matthews JP. Time
trends in accuracy of classification of testicular tumurs, with clinicalf and
epidemiological implications. Br J Cancer 1992;66:396-401.

Pearce N, Sheppard RA, Howard J K, Fraser J, Lilley BM. Time trends and

occupational differences in cancer of the testis in New Zealand. Cancer
1987;59: 1677-82.

124

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104.

105.

106.

Wilkinson T, Colls BM and Schluter PJ. Increased incidence of germ-cell
testicular cancer in New Zealand Maoris. Br J Cancer 1992;65:769-771.

Liu S, Semenciw R, Waters C, Wen SW, Mery LS, Mao Y. Clues to the
aetiological heterogeneity of testicular seminomas and non- seminomas: time
trends and age-period-cohort effects. Int J Epidemiol 2000;29:826-31.

Brown LM, Pottern LM, Hoover RN, Devesa SS, Aselton P, Flannery JT.
Testicular cancer in the United States: trends in incidence and mortality. Int J
Epidemiol 1986; 15: 164—70.

Zheng T, Holford TR, Ma Z, Ward BA, Flannery J, Boyle P. Continuing increase
in incidence of germ-cell testis cancer in young adults: experience from
Connecticut, USA, 1935-1992. Int J Cancer 1996;65:723-9.

Suominen J, Vierula M. Semen quality of Finnish men. ij 1993;306:1579.

Henderson BE, Benton B, J ing J, Yu MC, Pike MC. Risk factors for cancer of the
testis in young men. Int J Cancer 1979;23:598-602.

Cosgrove MD, Benton B, Henderson BE. Male genitourinary abnormalities and
maternal diethylstilbestrol. U Urol 1977;117:220-222.

Depue RH. Cryptorchidism, and epidemiologic study with emphasis on the
relationship to central nervous system dysfunction. Teratology 1988;37:301-5.

Giwercman A, Carlsen E, Keiding N, Skakkebaek NE. Evidence for increasing
incidence of abnormalities of the human testis: a review. Environ Health Perspect
1993;101 Suppl 2:65-71.

Carlsen E, Giwercman A, Skakkebaek NE. Declining sperm counts and
increasing incidence of testicular cancer and other gonadal disorders: is there a
connection? Ir Med J 1993;86:85-6.

James WH. Offspring sex ratio as a potential monitor of reproductive disorders in
communities near hazardous chemical sites [letter; comment] [see comments].
Reprod Toxicol 1997;] 1:893-5.

Giwercman A, Bonde JP. Declining male fertility and environmental factors.
Endocrinol Metab Clin North Am 1998;27:807-30, viii.

Skakkebaek NE, Berthelsen JG, Giwercman A, Muller J. Carcinoma-in-situ of the

testis: possible origin from gonocytes and precursor of all types of germ cell
tumours except spennatocytoma. Int J Androl 1987;10:19-28.

125

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

Potashnik G, Yanai-Inbar I. Dibromochloropropane (DBCP): an 8-year
reevaluation of testicular function and reproductive performance. Fertil Steril
1987;47:317-23.

J acobsen R, Bostofte E, Engholm G, et al. Risk of testicular cancer in men with
abnormal semen characteristics: cohort study. ij 2000;321:789-92.

Wilcox AJ, Baird DD, Weinberg CR, Homsby PP, Herbst AL. Fertility in men
exposed prenatally to diethylstilbestrol. N Engl J Med 1995;332:1411-6.

Strohsnitter WC, Noller KL, Hoover RN, et al. Cancer risk in men exposed in
utero to diethylstilbestrol. J Natl Cancer Inst 2001;93:545-51.

vom Saal FS, Timms BG, Montano MM, et a1. Prostate enlargement in mice due
to fetal exposure to low doses of estradiol or diethylstilbestrol and opposite effects
at high doses. Proc Natl Acad Sci U S A 1997;94:2056-61.

Safe SH. Endocrine disruptors and human health-—is there a problem? An update.
Environ Health Perspect 2000;108:487-93.

Weijin Z, Olsen J. Offspring sex ratio as an indicator of reproductive hazards
[letter; comment] [see comments]. Occup Environ Med 1996;53:503-4.

Lloyd OL, Fodor JG, Lloyd MM, Chockalingam A. The geography of the
community sex ratios for young children in Newfoundland. Sci Total Environ
1991;106:165-74.

Davis DL, Gottlieb MB, Stampnitzky JR. Reduced ratio of male to female births
in several industrial countries: a sentinel health indicator? [see comments]. Jama
1998;279: 1018-23.

Goldsmith JR. Dibromochloropropane: epidemiological findings and current
questions. Ann N Y Acad Sci 1997;837:300-6.

Marcus M, Kiely J, Xu F, McGeehin M, Jackson R, Sinks T. Changing sex ratio
in the United States, 1969-1995 [see comments]. Fertil Steril 1998;70:270-3.

Allan BB, Brant R, Seidel JE, Jarrell JF. Declining sex ratios in Canada [see
comments] [published erratum appears in Can Med Assoc J 1997 Feb
l;156(3):348]. Cmaj 1997;156:37-41.

Feitosa MF, Krieger H. Demography of the human sex ratio in some Latin
American countries, 1967-1986. Hum Biol 1992;64:523-30.

Moller H. Change in male:female ratio among newborn infants in Denmark
[letter; comment] [see comments]. Lancet 1996;348:828-9.

126

121.

122.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

van der Pal-de Bruin KM, Verloove-Vanhorick SP, Roeleveld N. Change in
malezfemale ratio among newborn babies in Netherlands [letter; comment] [see
comments]. Lancet 1997;349:62.

Bromen K, Jockel KH. Change in male proportion among newborn infants [letter;
comment]. Lancet 1997;349:804-5.

Dickinson HO, Parker L. Decline in sex ratios at birth, England and Wales, 1973-
90 [letter; comment]. J Epidemiol Community Health 1997;51:103.

Ohmi H, Hirooka K, Mochizuki Y. Reduced ratio of male to female births in
Japan [letter]. Int J Epidemiol 1999;28:597.

Ulizzi L, Zonta LA. Factors affecting the sex ratio in humans: multivariate
analysis of the Italian population. Hum Biol 1995;67:59-67.

Longnecker MP, Rogan WJ, Lucier G. The human health effects of DDT
(dichlorodiphenyltrichloroethane) and PCBS (polychlorinated biphenyls) and an

overview of organochlorines in public health. Annual Review Of Public Health
1997;18:211-44.

Safe S. Polychlorinated biphenyls (PCBs) and polybrominated biphenyls (PBBs):

biochemistry, toxicology, and mechanism of action. Crit Rev Toxicol
1984;13:319-95.

Giesy JP, Kannan K. Dioxin-like and non-dioxin-like toxic effects of
polychlorinated biphenyls (PCBS): implications for risk assessment. Crit Rev
Toxicol 1998;28:51 1-69.

McFarland VA, Clarke J U. Environmental occurrence, abundance, and potential
toxicity of polychlorinated biphenyl congeners: considerations for a congener-
specific analysis. Environ Health Perspect 1989;81:225-39.

Smith S, Schecter A, Papke 0, Do T, Coulibaly D, Brandt-Rauf P. Quantitation of
the extracellular domain of epidermal growth factor receptor in the plasma of
dioxin-exposed individuals. Am J Ind Med 1998;34:1-5.

Tian Y, Ke S, Denison MS, Rabson AB, Gallo MA. Ah receptor and NF—kappa B
interactions, a potential mechanism for dioxin toxicity. Journal Of Biological
Chemistry 1999;274:510-5.

Kafafi SA, Afeefy HY, Ali AH, Said HK, Kafafi AG. Binding of polychlorinated

biphenyls to the aryl hydrocarbon receptor. Environ Health Perspect
1993;101:422-8.

127

133.

134.

137.

138.

139.

140.

141.

142.

143.

144.

Landers JP, Bunce NJ. The Ah receptor and the mechanism of dioxin toxicity.
Biochemical Journal 1991;276 ( Pt 2):273-87.

Safe S. Polychlorinated biphenyls (PCBs), dibenzo-p-dioxins (PCDDs),
dibenzofurans (PCDFs), and related compounds: environmental and mechanistic

considerations which support the development of toxic equivalency factors
(TEFs). Crit Rev Toxicol 1990;21:51-88.

Bitman J, Cecil HC. Estrogenic activity of DDT analogs and polychlorinated
biphenyls. J Agric Food Chem 1970;18:1108-12.

Safe S, Astroff B, Harris M, et a1. 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)
and related compounds as antioestrogens: characterization and mechanism of
action. Pharmacol Toxicol 1991;69:400-9.

Buchanan DL, Sato T, Peterson RE, Cooke PS. Antiestrogenic effects of 2,3,7,8-
tetrachlorodibenzo-p—dioxin in mouse uterus: critical role of the aryl hydrocarbon
receptor in stromal tissue. Toxicol Sci 2000;57:302-11.

Seegal RF. Epidemiological and laboratory evidence of PCB-induced
neurotoxicity. Crit Rev Toxicol 1996;26:709-37.

Fischer LJ, Seegal RF, Ganey PE, Pessah 1N, Kodavanti PR. Symposium
overview: toxicity of non-coplanar PCBS. Toxicol Sci 1998;41:49-61.

Schantz SL, Moshtaghian J, Ness DK. Spatial learning deficits in adult rats
exposed to ortho-substituted PCB congeners during gestation and lactation.
Fundam Appl Toxicol 1995;26:1 17-26.

Schantz SL, Seo BW, Wong PW, Pessah IN. Long-term effects of developmental
exposure to 2,2',3,5',6- pentachlorobiphenyl (PCB 95) on locomotor activity,

spatial learning and memory and brain ryanodine binding. Neurotoxicology
1997;18:457-67.

Jacobson JL, Jacobson SW. Intellectual impairment in children exposed to
polychlorinated biphenyls in utero. N Engl J Med 1996;335:783-9.

Schantz SL, Gardiner J C, Gasior DM, Sweeney AM, Humphrey HE, McCaffrey
RJ. Motor function in aging Great Lakes fisheaters. Environ Res 1999;80:846-
SS6.

Tithof PK, Watts S, Ganey PE. Protein tyrosine kinase involvement in the
production of superoxide anion by neutrophils exposed to Aroclor 1242, a mixture
of polychlorinated biphenyls. Biochem Pharmacol 1997;53:1833—42.

128

145.

146.

147.

148.

149.

150.

151.

152.

153.

154.

155.

156.

Humphrey HE, Gardiner J C, Pandya JR, et al. PCB congener profile in the serum
of humans consuming great lakes fish [In Process Citation]. Environ Health
Perspect 2000;108: 167-72.

DeVoto E, Fiore BJ, Millikan R, et al. Correlations among human blood levels of
specific PCB congeners and implications for epidemiologic studies. Am J Ind
Med 1997;32:606-13.

Korrick SA, Altshul L. High breast milk levels of polychlorinated biphenyls
(PCBs) among four women living adjacent to a PCB-contaminated waste site.
Environ Health Perspect 1998;106:513-8.

Van Birgelen A, DeVito MJ and Bimbaum LS. Toxic equivalency factors derived
from cytochrome P450 induction in mice are predictive for cytochrome P450
induction after subchronic exposure to mixtures of PCDDs, PCDFs, and PCBS in
female B6C3F1 mice and Sprague Dawley rats. Organohalogen Compounds
1996;29:251-256.

Vonier PM, Crain DA, McLachlan JA, Guillette LJ, Arnold SF. Interaction of
environmental chemicals with the estrogen and progesterone receptors from the
oviduct of the American alligator. Environ Health Perspect 1996;104: 1318-22.

Arnold SF, Klotz DM, Collins BM, Vonier PM, Guillette LJ, McLachlan JA.
Synergistic activation of estrogen receptor with combinations of environmental
chemicals. Science 1996;272:1489-92.

Arnold SF, Bergeron J M, Tran DQ, et a1. Synergistic responses of steroidal
estrogens in vitro (yeast) and in vivo (turtles). Biochem Biophys Res Commun
1997;235:336-42.

Stow CA, Carpenter SR, Eby LA, Amrhein JF, Hesselberg RJ. Evidence that
PCBs are approaching stable concentrations in Lake Michigan fishes. Ecological
Applications 1995;52248-260.

Fensterheim RJ. Documenting temporal trends of polychlorinated biphenyls in the
environment. Regul. Toxicol. Pharmacol 1993;18:181-201.
Agency UEP. Fed Regist 1993;58:20806-09.

Webber T. Study says banned toxin is entering Lake Michigan. Lansing State
Journal. Lansing, MI, 2000.

Wong C ea. Accumulation and preliminary inventory of PCBs and organochlorine
pesticides in Lake Ontario sediments. First international conference on
contaminated sedimentszhistorical records, environmental impact and
remediation. Milwaukee, WI, 1993.

129

157.

158.

159.

160.

161.

162.

163.

164.

165.

166.

167.

168.

169.

170.

Provinces NSECP. Mercury Study. A framework for action, 1998.

Harris RaH, C. Mercury-measuring and managing the risk. Environment
1978;20:25-36.

Wilkinson S, Golding, KH and Robinson, PK. Mercury accumulation and
volatilization in immobilzed agal cell systems. Biotechnol Lett 1989;] 1:861-864.

EPA. The USEPA Mercury Study: Report to Congress. Volume I: Executive
summary. 1997.

Fitzgerald WF, Clarkson TW. Mercury and monomethylmercury: present and
future concerns. Environ Health Perspect 1991;96:159-66.

EPA. Mercury Study: Report to Congress. Volume II]: An assessment of
exposure from anthropogenic mercury emmissions in the US: USEPA, 1996.

Michigan. Michigan fish contaminant monitoring program. Lansing: State of
Michigan, 1998:8-9.

Stevens J aB, BE. Occurrence of heavy metals in the blue shark, Prionace glauca,
and selected pelagic fish in the north-east Atlantic Ocean. Mar Biol 1974;26:287-
293.

Galal-Gorchev H. Dietary intake, levels in food and estimated intake of lead,
cadmium, and mercury. Food Addit Contam 1993;10:115-28.

Organization WHO. Mercury. Geneva, Switzerland: WHO, 1989:] 15.

Day R. Michigan Fish Contaminant Monitoring Program 1998 Annual Report:
Michigan Department of Environmental Quality-Surface Water Quality Division,
1998.

Zabik ME, Zabik MJ, Booren AM, et a1. Pesticides and total polychlorinated
biphenyls residues in raw and cooked walleye and white bass harvested from the
Great Lakes. Bull Environ Contam Toxicol 1995;54:396-402.

DeVault DS, Hesselberg R, Rodgers PW, Feist TJ. Contaminant Trends in Lake
Trout and Walleye from the Laurentian Great Lakes. J. Great lakes Res.
1996;22:884-895.

Kreiss K. Studies on populations exposed to polychlorinated biphenyls. Environ
Health Perspect 1985;60:193-9.

130

171.

172.

173.

174.

175.

176.

177.

178.

179.

180.

181.

182.

Hanrahan LP, Falk C, Anderson HA, Draheim L, Kanarek MS, Olson J. Serum
PCB and DDE levels of frequent Great Lakes sport fish consumers-a first look.
The Great Lakes Consortium. Environ Res 1999;80:S26-S37.

Hanrahan LP FC, Anderson HA, Draheim L, Steeprot D, Olson J, Fiore B,
Kanarek M, and the Great Lakes Consortium. Serum PCB Levels and Great Lakes
Sport Fish Consumption. Health Conference '97 Great Lakes and St. Lawrence.
Montreal Quebec, Canada, 1997.

Falk C, Hanrahan L, Anderson HA, et a]. Body burden levels of dioxin, furans,
and PCBS among frequent consumers of Great Lakes sport fish. The Great Lakes
Consortium. Environ Res 1999;80:S19-825.

Hovinga ME SM, and Humphrey HEB. Historical changes in serum PCB and
DDT levels in an environmentally-exposed cohort. Arch. Environ. Contam.
Toxicol. 1992;22:362-366.

Resources MDNR. Resources and conservation: MI DNR, 2000.

Johnson BL HH, De Rosa T. Key Environmental Human Health Issues in the
Great Lakes and St. Lawrence River Basins. Environ. Research Section A

1999;80:82-812.

He J -P, Humphrey HEB, Paneth N, Courval JM. Time trends in sport—caught

Great Lakes fish consumption and serum polychlorinated biphenyl levels among
Michigan anglers, 1973-1999. Environ Sci Technol 2001;35:435-440.

Humphrey HE, Budd ML. Michigan's fisheater cohorts: a prospective history of
exposure. Toxicol Ind Health 1996;12:499-505.

Kearney JP, Cole DC, Ferron LA, Weber JP. Blood PCB, p,p'-DDE, and mirex
levels in Great Lakes fish and waterfowl consumers in two Ontario communities.
Environ Res 1999;80:S138-Sl49.

Robinson PE, Mack GA, Remmers J, Levy R, Mohadjer L. Trends of PCB,
hexachlorobenzene, and beta-benzene hexachloride levels in the adipose tissue of

the US. population. Environ Res 1990;53:175-92.

Airey D. Total mercury concentrations in human hair from 13 countries in
relation to fish consumption and location. Sci Total Environ 1983;31:157-80.

Airey D. Mercury in human hair due to environment and diet: a review. Environ
Health Perspect 1983;52:303-16.

Stern AH. Re-evaluation of the reference dose for methylmercury and assessment
of current exposure levels. Risk Anal 1993;13:355-64.

13]

184.

185.

186.

187.

188.

189.

190.

191.

192.

193.

194.

195.

Clarkson T. An outbreak of mercury poisoning due to consumption of
contaminated grain. Fed Proc 1975;34:2395-2399.

Gerstenberger SL, Tavris DR, Hansen LK, Pratt-Shelley J, Dellinger JA.
Concentrations of blood and hair mercury and serum PCBs in an Ojibwa

population that consumes Great Lakes region fish. J Toxicol Clin Toxicol
1997;35:377-86.

WHO. Methylmercury. Geneva, Switzerland: World Health Organization, 1990.

Peterson RE, Theobald HM, Kimmel GL. Developmental and reproductive
toxicity of dioxins and related compounds: cross-species comparisons. Crit Rev
Toxicol 1993;23:283-335.

Tillet DE AG, and Giesy JP. Planar chlorinated hydrocarbons (PCH's) in colonial
fish-eating waterbirds eggs from the Great Lakes. Govt. reports Announcements
and Index (GRS&I) 199].

TR. C. Toxicants in the Aquaeous Ecosystem: John Wiley & Sons Ltd., 1979.

Mac MaS, TR. Investigations into the effects of PCB congeners on reproduction
in lake trout from the Great Lakes. Chemosphere 1992;25:189-192.

Jennings M, Percival HF, and Woodward, AR. Evaluation of alligator hatchlings
and egg removal from three Florida lakes. Proc Ann Conf Southeast Assoc Fish
Wildl Agencies 1988;42:283-294.

Guillette LI, Jr., Pickford DB, Crain DA, Rooney AA, Percival HF. Reduction in
penis size and plasma testosterone concentrations in juvenile alligators living in a
contaminated environment. Gen Comp Endocrinol 1996;101:32-42.

Guillette LI, Gross TS, Masson GR, Matter J M, Percival HF, Woodward AR.
Developmental abnormalities of the gonad and abnormal sex hormone

concentrations in juvenile alligators from contaminated and control lakes in
Florida. Environ Health Perspect 1994;102:680-8.

Guillette LJ, Gross TS, Gross DA, Rooney AA, Percival HF. Gonadal
steroidogenesis in vitro from juvenile alligators obtained from contaminated or
control lakes. Environ Health Perspect 1995;103 Suppl 4:31-6.

Roelke ME SD, Facemire CF and Sundlof SF. Mercury contamination in the free-

ranging endangered Florida panter (Felis concolor coryi). Am Assoc Zoo Vet
Annu Proc 1990;1991:277-283.

132

196.

197.

198.

I99.

200.

201.

202.

203.

204.

205.

206.

207.

Roelke ME MJaOBS. The consequences of demographhic reduction and genetic
depletion in the endangered Florida panther. Curr Biol 1993;3z340-350.

Facemire CF, Gross TS, Guillette Ll, Jr. Reproductive impairment in the Florida
panther: nature or nurture? Environ Health Perspect 1995;103 Suppl 4:79-86.

Guillette LI, Crain DA, Rooney AA, Pickford DB. Organization versus
activation: the role of endocrine-disrupting contaminants (EDCs) during
embryonic development in wildlife. Environ Health Perspect 1995;103 Suppl
72157-64.

Aulerich RJ, Ringer RK. Current status of PCB toxicity to mink, and effect on
their reproduction. Arch Environ Contam Toxicol 1977;62279-92.

Hornshaw TC, Aulerich RJ, Johnson HE. Feeding Great Lakes fish to mink:
effects on mink and accumulation and elimination of PCBS by mink. J Toxicol
Environ Health 1983;] 1:933-46.

Allen JR, Barsotti DA, Lambrecht LK, Van Miller JP. Reproductive effects of
halogenated aromatic hydrocarbons on nonhuman primates. Ann N Y Acad Sci
1979;320:419-25.

Sager DB. Effect of postnatal exposure to polychlorinated biphenyls on adult
male reproductive function. Environ Res 1983;31:76-94.

Johansson B. Lack of effects of polychlorinated biphenyls on testosterone
synthesis in mice. Pharmacol Toxicol 1987;61:220-3.

Sanders OT, Kirkpatrick RL, Scanlon PE. Polychlorinated biphenyls and
nutritional restriction: their effects and interactions on endocrine and reproductive
characteristics of male white mice. Toxicol Appl Pharmacol 1977;40:91-8.

Moore RW, Potter CL, Theobald HM, Robinson J A, Peterson RE. Androgenic
deficiency in male rats treated with 2,3,7,8- tetrachlorodibenzo-p-dioxin. Toxicol
Appl Pharmacol 1985;79:99-1 11.

Moore RW, Jefcoate CR, Peterson RE. 2,3,7,8-Tetrachlorodibenzo-p-dioxin
inhibits steroidogenesis in the rat testis by inhibiting the mobilization of
cholesterol to cytochrome P4508cc. Toxicol Appl Pharmacol 1991;109:85-97.

Faqi AS, Dalsenter PR, Mathar W, Heinrich-Hirsch B, Chahoud 1. Reproductive

toxicity and tissue concentrations of 3,3’,4,4'- tetrachlorobiphenyl (PCB 77) in
male adult rats. Hum Exp Toxicol 1998; 17: 151-6.

133

208.

209.

210.

211.

212.

213.

214.

215.

216.

217.

218.

219.

Humphrey HEB. Chemical contaminants in the Great Lakes: The human health
aspect. In: M E, ed. Toxic Contaminants and Ecosystem Health: A Great Lakes
Focus: Wiley, New York, 1988:153-165.

DeVito MJ, Bimbaum LS, Farland WH, Gasiewicz TA. Comparisons of
estimated human body burdens of dioxinlike chemicals and TCDD body burdens
in experimentally exposed animals. Environ Health Perspect 1995;103:820-31.

Roman E, and Peterson, RE. Developmental male reproductive toxicology of
2,3,7,8—tetrachlorodibenzo-p-dioxin (TCDD) and PCBS. In: Korach, KS, ed.
Reproductive and developmental toxicology, New York: Marcel Dekker, 1998:
593-624. 1998.

Mocarelli P GP, Ferrari E, et al. Paternal concentrations of dioxin and sex ratio of
offspring. Lancet 2000;355:1858-1863.

Styne D. Puberty. In: GJ GFaS, ed. Basic and Clinical Endocrinology. Stamford,
CT: Appleton & Lange, 1997:521-547.

Huang A, Powell D, Chou K. Pre- and postnatal exposure to 3,3',4,4'-
tetrachlorobiphenyl: I. Effects on breeding ability and sperm fertilizing ability in
male mice. Arch Environ Contam Toxicol 1998;34:204-8.

Sager DB, Shih-Schroeder W, Girard D. Effect of early postnatal exposure to
polychlorinated biphenyls (PCBS) on fertility in male rats. Bull Environ Contam
Toxicol 1987;38:946-53.

Shain W, Ovennann SR, Wilson LR, Kostas J, Bush B. A congener analysis of
polychlorinated biphenyls accumulating in rat pups after perinatal exposure. Arch
Environ Contam Toxicol 1986;15:687-707.

Sager D, Girar, D and Nelson, D. Early postnatal exposure to PCBs: sperm
function in rats. Environ Toxicol Chem 1991;10:737-746.

Bush B, Bennett AH, Snow JT. Polychlorobiphenyl congeners, p,p'-DDE, and
sperm function in humans. Arch Environ Contam Toxicol 1986;15:333-41.

Hauser R, Altshul L, Schiff I, et al. Semen quality and polychlorinated biphenyls.
Recent Advances in the Environmental Toxicology and Health Effects of EPCs
(PCB Workshop) 2000;2000.

Courval J DJ, Stein A, Tay E, He J, Humphrey H, Paneth N. Sport-Caught Fish

Consumption and Conception Delay in Liscensed Michigan Anglers.
Environmental Research Section A 1999;80:S 183-3188.

134

 

220.

221.

222.

223.

224.

225.

226.

227.

228.

229.

230.

231.

232.

Buck GM, Mendola P, Vena J E, et a1. Paternal Lake Ontario fish consumption
and risk of conception delay, New York State Angler Cohort. Environmental
Research 1999;80:813-818.

Emmett EA, Maroni M, Jefferys J, Schmith J, Levin BK, Alvares A. Studies of
transformer repair workers exposed to PCBs: II. Results of clinical laboratory
investigations. Am J Ind Med 1988;14:47-62.

Dougherty R, Whitaker, MJ, Tang, SY et a1. Sperm density and toxic substances:
a potential key to environmental health hazards. Environ Health Chem
1980;13:268.

Schlebusch H, Wagner U, van der Ven H, al-Hasani S, Diedrich K, Krebs D.
Polychlorinated biphenyls: the occurrence of the main congeners in follicular and
sperm fluids. J Clin Chem Clin Biochem 1989;27:663—7.

Guo YL, Hsu PC, Hsu CC, Lambert GH. Semen quality after prenatal exposure to
polychlorinated biphenyls and dibenzofurans. Lancet 2000;356:1240-1.

Whorton D, Krauss RM, Marshall S, Milby TH. Infertility in male pesticide
workers. Lancet 1977;2: 1259-61.

Whorton D, Milby TH, Krauss RM, Stubbs HA. Testicular function in DBCP
exposed pesticide workers. J Occup Med 1979;21:161-6.

Potashnik G, Ben-Aderet N, Israeli R, Yanai-Inbar I, Sober I. Suppressive effect

of l,2-dibromo-3-chloropropane on human spermatogenesis. Fertil Steril
1978;30:444-7.

Potashnik G, Goldsmith J, Insler V. Dibromochloropropane—induced reduction of
the sex-ratio in man. Andrologia 1984;16:213-8.

Goldsmith JR, Potashnik G, Israeli R. Reproductive outcomes in families of
DBCP-exposed men. Arch Environ Health 1984;39:85-9.

Glass R, Lyness, RN, Mengle, DC, et a1. Sperm count depression in pesticide
applicators exposed to dibromochloropropane. Am J Epidemiology

1979;109:346-351.

Potashnik G. A four-year reassessment of workers with dibromochloropropane-
induced testicular dysfunction. Andrologia 1983;15:164-70.

Curtis KM SD, Weinberg CR and Arbuckle TE. The effect of pesticide exposure
on time to pregnancy. Epidemiology 1999;10:112-117.

135

233.

234.

235.

236.

237.

238.

239.

240.

241.

242.

243.

244.

Savitz DA, Arbuckle T, Kaczor D, Curtis KM. Male pesticide exposure and
pregnancy outcome. Am J Epidemiol 1997;146:1025-36.

Thonneau P, Abel] A, Larsen SB, et al. Effects of pesticide exposure on time to
pregnancy: results of a multicenter study in France and Denmark. ASCLEPIOS
Study Group. Am J Epidemiol 1999;150:157-63.

Sharpe RM, Skakkebaek NE. Are oestrogens involved in falling sperm counts and
disorders of the male reproductive tract? [see comments]. Lancet 1993;341:1392-
5.

Kelce WR, Monosson B, Gray LE, Jr. An environmental antiandrogen. Recent
Prog Horm Res 1995;50:449-53.

Akingbemi BT, Ge RS, Klinefelter GR, Gunsalus GL, Hardy MP. A metabolite of
methoxychlor, 2,2-bis(p-hydroxypheny1)-1,1, l- trichloroethane, reduces
testosterone biosynthesis in rat leydig cells through suppression of steady-state

messenger ribonucleic acid levels of the cholesterol side-chain cleavage enzyme.
Biol Reprod 2000;62:571—8.

Waliszewski SM, Szymczynski GA. Determination of selected chlorinated
pesticides, bound and free, in human semen. Arch Environ Contam Toxicol
1983;12:577-80.

Kociba RJ, Keeler PA, Park CN, Gehring PJ. 2,3,7,8-tetrachlorodibenzo-p-dioxin
(TCDD): results of a 13-week oral toxicity study in rats. Toxicol Appl Pharmacol
1976;35:553-74.

McConnell BB, Moore J A, Haseman J K, Harris MW. The comparative toxicity of
chlorinated dibenzo-p-dioxins in mice and guinea pigs. Toxicol Appl Pharmacol
1978;44:335-56.

Moore J A HM, AP. Tissue distribution of [14C]tetrachlorodibenzo-p-dioxin in
pregnant and neonatal rats. Toxicol Appl Pharmacol 1976;37:146-147.

Mebus CA, Reddy VR, Piper WN. Depression of rat testicular l7-hydroxylase

and 17,20-1yase after administration of 2,3,7,8—tetrachlorodibenzo-p-dioxin
(TCDD). Biochem Pharmacol 1987;36:727-31.

Chahoud I, Krowke R, Schimmel A, Merker HJ, Neubert D. Reproductive
toxicity and pharrnacokinetics of 2,3,7,8- tetrachlorodibenzo-p-dioxin. 1. Effects

of high doses on the fertility of male rats. Arch Toxicol 1989;63:432-9.

Kleeman JM, Moore RW, Peterson RE. Inhibition of testicular steroidogenesis in
2,3,7,8-tetrachlorodibenzo- p—dioxin-treated rats: evidence that the key lesion

136

245.

246.

247.

248.

249.

250.

251.

252.

253.

254.

occurs prior to or during pregnenolone formation. Toxicol Appl Pharmacol
1990;106:112-25.

Mably TA, Moore RW, Peterson RE. In utero and lactational exposure of male
rats to 2,3,7,8- tetrachlorodibenzo-p-dioxin. 1. Effects on androgenic status.
Toxicol Appl Pharmacol 1992;l 14:97-107.

Mably TA, Moore RW, Goy RW, Peterson RE. In utero and lactational exposure
of male rats to 2,3,7,8- tetrachlorodibenzo-p-dioxin. 2. Effects on sexual behavior

and the regulation of luteinizing hormone secretion in adulthood. Toxicol Appl
Pharmacol 1992;] 14: 108-17.

Mably TA, Bjerke DL, Moore RW, Gendron-Fitzpatrick A, Peterson RE. In utero
and lactational exposure of male rats to 2,3,7,8- tetrachlorodibenzo-p-dioxin. 3.

Effects on spermatogenesis and reproductive capability. Toxicol Appl Pharmacol
1992;114:118-26.

Faqi AS, Dalsenter PR, Merker HJ, Chahoud I. Reproductive toxicity and tissue
concentrations of low doses of 2,3,7,8- tetrachlorodibenzo-p—dioxin in male

offspring rats exposed throughout pregnancy and lactation. Toxicol Appl
Pharmacol 1998;150:383-92.

Gray LE, Ostby J S, Kelce WR. A dose-response analysis of the reproductive
effects of a single gestational dose of 2,3,7,8-tetrachlorodibenzo-p—dioxin in male
Long Evans Hooded rat offspring. Toxicol Appl Pharmacol 1997;146:11-20.

Egeland GM, Sweeney MH, Fingerhut MA, Wille KK, Schnorr TM, Halperin
WE. Total serum testosterone and gonadotropins in workers exposed to dioxin
[see comments]. Am J Epidemiol 1994;139:272-81.

Rachootin P, Olsen J. The risk of infertility and delayed conception associated
with exposures in the Danish workplace. J Occup Med 1983;25:394-402.

Schecter A, McGee H, Stanley J S, Boggess K, Brandt-Rauf P. Dioxins and
dioxin-like chemicals in blood and semen of American Vietnam veterans from the
state of Michigan [published erratum appears in Am J Ind Med 1997
Mar;31(3):370-1]. Am J Ind Med 1996;30:647-54.

De Kretser D. Morphology and physiology of the testis. In: Becker K, ed.
Principles and practice of endocrinology and metabolism. Philadelphia, PA: J.B.
Lippincott company, 1995: 1032-1041 .

Kuratsune M YT, Matsuzaka J and Yamaguchi A. Epidemilogical study on

Yusho, a poisoning caused by ingestion of rice oil contaminated with commercial
brand of polychlorinated biphenyls. Environ Hlth Perspec l972;1:119-128.

137

255.

256.

257.

258.

259.

260.

261.

262.

263.

264.

265.

266.

Hsu ST, Ma CI, Hsu SK, Wu SS, Hsu NH, Yeh CC. Discovery and epidemiology
of PCB poisoning in Taiwan. Prog Clin Biol Res 1984;137:71-9.

Rogan WJ, Gladen BC, Guo YL, Hsu CC. Sex ratio after exposure to dioxin-like
chemicals in Taiwan. Lancet 1999;353:206-7.

Arnold DL, Bryce F, McGuire PF, et a1. Toxicological consequences of aroclor
1254 ingestion by female rhesus (Macaca mulatta) monkeys. Part 2. Reproduction
and infant findings [see comments]. Food Chem Toxicol 1995;33:457-74.

Simmons DL, Valentine DM, Bradshaw WS. Different patterns of developmental
toxicity in the rat following prenatal administration of structurally diverse
chemicals. J Toxicol Environ Health 1984;14:121-36.

Mocarelli P, Needham LL, Marocchi A, et a1. Serum concentrations of 2,3,7,8-
tetrachlorodibenzo-p-dioxin and test results from selected residents of Seveso,
Italy. J Toxicol Environ Health 1991;32:357-66.

Mocarelli P, Brambilla P, Gerthoux PM, Patterson DG, Jr., Needham LL. Change
in sex ratio with exposure to dioxin [letter]. Lancet 1996;348:409.

Nagel SC vSFTCea. Relative binding affinity-serum modified access (RBA-
SMA) assay predicts the relative in vivo bioactivity of the xenoestrogens
bisphenol A and octyphenol. Environ Health Perspect 1997;105:70-77.

Young AL CJ, Thalken CE and Trembly JW. The toxicology, environmental fate
and human risk of herbicide orange and its associated dioxin. Brooks Air Force
Base, TX: Medical Division: USAF Occupational and Environmental Health
Laboratory (OEHL), 1978.

Stellman SD, Stellman J M, Sommer JF, Jr. Combat and herbicide exposures in
Vietnam among a sample of American Legionnaires. Environ Res 1988;47:112-
28.

Potashnik G, Porath A. Dibromochloropropane (DBCP): a l7-year reassessment
of testicular function and reproductive performance. J Occup Environ Med
1995;37:1287-92.

Whorton MD, Wong O, Morgan RW, Gordon N. An epidemiologic investigation
of birth outcomes in relation to dibromochloropropane contamination in drinking
water in Fresno County, California, USA. Int Arch Occup Environ Health
1989;61:403-7.

Reich MR, Spong JK. Kepone: a chemical disaster in Hopewell, Virginia. Int J
Health Serv 1983;13:227-46.

138

267.

268.

269.

270.

271.

272.

273.

274.

275.

276.

277.

278.

Cannon SB, Veazey JM, Jr., Jackson RS, et al. Epidemic kepone poisoning in
chemical workers. Am J Epidemiol 1978;107:529-37.

Taylor JR, Selhorst JB, Houff SA, Martinez AJ. Chlordecone intoxication in man.
I. Clinical observations. Neurology 1978;28:626-30.

Silberhorn EM, Glauert HP, Robertson LW. Carcinogenicity of polyhalogenated
biphenyls: PCBs and PBBs. Crit Rev Toxicol 1990;20:440-96.

Parkinson A, Safe SH, Robertson LW, et al. Immunochemical quantitation of
cytochrome P-450 isozymes and epoxide hydrolase in liver microsomes from
polychlorinated or polybrominated biphenyl-treated rats. A study of structure-
activity relationships. J Biol Chem 1983;258:5967-76.

Shimada T, Yun CH, Yamazaki H, Gautier JC, Beaune PH, Guengerich FP.
Characterization of human lung microsomal cytochrome P-450 1A1 and its role in
the oxidation of chemical carcinogens. Mol Pharmacol 1992;41:856-64.

Shimada T, Hayes CL, Yamazaki H, et al. Activation of chemically diverse
procarcinogens by human cytochrome P- 450 1B1. Cancer Res 1996;56:2979-84.

Pang S, Cao, JQ, Katz, BH, et a1. Inductive and inhibitory effects of non-ortho-

substituted polychlorinated biphenyls on estrogen metabolism and human
cytochromes P450 1A1 and 1B1. Biochem Pharmacol 1999;58:29-38.

Kawajiri KaHS-I. The CYPl Family. In: C 1, ed. Cytochromes P450 Metabolic
and Toxicological Aspects: CRC Press, Inc, 1996:77-97.

Masten SA G], Miller CR et al. Population-based studies of dioxin
responsiveness: individual variation in CYP1A1 levels and relationship to dioxin
body burden. Organohalogen Compounds 1998;38:13-16.

Spink DC, Hayes CL, Young NR, et al. The effects of 2,3,7,8-tetrachlorodibenzo-
p-dioxin on estrogen metabolism in MCF-7 breast cancer cells: evidence for

induction of a novel 17 beta-estradiol 4-hydroxylase. J Steroid Biochem Mol Biol
1994;51:251-8.

Zhu BT, Conney AH. Functional role of estrogen metabolism in target cells:
review and perspectives. Carcinogenesis 1998;19:1-27.

Sutter TR, Tang YM, Hayes CL, et al. Complete cDNA sequence of a human

dioxin-inducible mRN A identifies a new gene subfamily of cytochrome P450 that
maps to chromosome 2. J Biol Chem 1994;269:13092-9.

139

279.

280.

281.

282.

283.

284.

285.

286.

287.

288.

289.

290.

Liehr JG, Ricci MJ. 4-Hydroxylation of estrogens as marker of human mammary
tumors. Proc Natl Acad Sci U S A 1996;93:3294—6.

Han X, Liehr JG. 8-Hydroxylation of guanine bases in kidney and liver DNA of
hamsters treated with estradiol: role of free radicals in estrogen-induced
carcinogenesis. Cancer Res 1994;54:5515-7.

Worrnhoudt LW CJaNP. Genetic polymorphisms of human N-acetyltransferase,
cytochrome P450, glutathione-S-transferase, and epoxide hydrolase enzymes:
relevance to xenobiotic metabolism and toxicity. Crit Rev Toxicol 1999;29:59-
124.

Moysich KB, Shields PG, Freudenheim JL, et al. Polychlorinated biphenyls,
cytochrome P4501A1 polymorphism, and postmenopausal breast cancer risk.
Cancer Epidemiol Biomarkers Prev 1999;8241-4.

Shields PG, Caporaso NE, Falk RT, et a1. Lung cancer, race, and a CYP1A1
genetic polymorphism. Cancer Epidemiol Biomarkers Prev 1993;2:481-5.

Zhan g ZY, Fasco MJ, Huang L, Guengerich FP, Kaminsky LS. Characterization
of purified human recombinant cytochrome P4501A1- Ile462 and -Val462:
assessment of a role for the rare allele in carcinogenesis. Cancer Res
1996;56:3926-33.

Cosma G, Crofts F, Taioli E, Toniolo P, Garte S. Relationship between genotype
and function of the human CYP1A1 gene. J Toxicol Environ Health 1993;40:309-
16.

Ambrosone CB, Freudenheim JL, Graham S, et al. Cytochrome P4501A1 and
glutathione S-transferase (M1) genetic polymorphisms and postmenopausal breast
cancer risk. Cancer Res 1995;55:3483-5.

Ishibe N, Hankinson SE, Colditz GA, et a1. Cigarette smoking, cytochrome P450

1A1 polymorphisms, and breast cancer risk in the Nurses' Health Study. Cancer
Res 1998;58:667-71.

Baron J A. Smoking and estrogen-related disease. Am J Epidemiol 1984;119:9-22.
Taioli E, Trachman J, Chen X, Toniolo P, Garte SJ. A CYP1A1 restriction
fragment length polymorphism is associated with breast cancer in African-
American women. Cancer Res 1995;55:3757-8.

Jellinck PH, Forkert PG, Riddick DS, Okey AB, Michnovicz JJ, Bradlow HL. Ah

receptor binding properties of indole carbinols and induction of hepatic estradiol
hydroxylation. Biochem Pharmacol 1993;45: 1 129-36.

140

291.

292.

293.

294.

295.

296.

297.

298.

299.

300.

Oakley GG, Robertson LW, Gupta RC. Analysis of polychlorinated biphenyl-
DNA adducts by 32P-post1abeling. Carcinogenesis 1996;17:109-14.

Amaro AR, Oakley GG, Bauer U, Spielmann HP, Robertson LW. Metabolic
activation of PCBs to quinones: reactivity toward nitrogen and sulfur nucleophiles
and influence of superoxide dismutase. Chem Res Toxicol 1996;9:623-9.

Dalet-Beluche I, Boulenc X, Fabre G, Maurel P, Bonfils C. Purification of two
cytochrome P450 isozymes related to CYP2A and CYP3A gene families from

monkey (baboon, Papio papio) liver microsomes. Cross reactivity with human
forms. Eur J Biochem 1992;204:641-8.

Rebbeck TR, J affe JM, Walker AH, Wein AJ, Malkowicz SB. Modification of
clinical presentation of prostate tumors by a novel genetic variant in CYP3A4
[published erratum appears in J Natl Cancer Inst 1999 Jun 16;91(12):1082] [see
comments]. J Natl Cancer Inst 1998;90:1225-9.

Golden RJ, Noller KL, Titus-Ernstoff L, et al. Environmental endocrine
modulators and human health: an assessment of the biological evidence. Crit Rev
Toxicol 1998;28:109-227.

Danzo BJ. Environmental xenobiotics may disrupt normal endocrine function by
interfering with the binding of physiological ligands to steroid receptors and
binding proteins. Environ Health Perspect 1997;105:294-301.

Roman BL, Sommer RJ, Shinomiya K, Peterson RE. In utero and lactational
exposure of the male rat to 2,3,7,8- tetrachlorodibenzo-p-dioxin: impaired prostate
growth and development without inhibited androgen production. Toxicol Appl
Pharmacol 1995;134:241-50.

Krishnan V, Safe S. Polychlorinated biphenyls (PCBs), dibenzo—p-dioxins
(PCDDs), and dibenzofurans (PCDFs) as antiestrogens in MCF—7 human breast
cancer cells: quantitative structure—activity relationships. Toxicol Appl Pharmacol
1993;120:55-61.

Yeowell HN, Waxman DJ, LeBlanc GA, Linko P, Goldstein JA. Suppression of
male-specific cytochrome P450 2c and its mRNA by 3,4,5,3',4',5'-
hexachlorobiphenyl in rat liver is not causally related to changes in serum
testosterone. Arch Biochem Biophys 1989;271:508-14.

Bjerke DL, Brown TJ, MacLusky NJ, Hochberg RB, Peterson RE. Partial
demasculinization and feminization of sex behavior in male rats by in utero and
lactational exposure to 2,3,7,8-tetrachlorodibenzo-p- dioxin is not associated with

alterations in estrogen receptor binding or volumes of sexually differentiated brain
nuclei. Toxicol Appl Pharmacol 1994;127:258-67.

141

301.

303.

304.

305.

306.

307.

308.

309.

310.

311.

Bergeron J M, Crews D, McLachlan J A. PCBS as environmental estrogens: turtle
sex determination as a biomarker of environmental contamination. Environ Health
Perspect 1994; 102:780- 1.

Commandeur JN, Stijntjes GJ, Vermeulen NP. Enzymes and transport systems
involved in the formation and disposition of glutathione S-conjugates. Role in
bioactivation and detoxication mechanisms of xenobiotics. Pharmacol Rev
1995;47:271-330.

Seidegard J, Vorachek WR, Pero RW, Pearson WR. Hereditary differences in the
expression of the human glutathione transferase active on trans-stilbene oxide are
due to a gene deletion. Proc Natl Acad Sci U S A 1988;85:7293-7.

McWilliams JE, Sanderson BJ, Harris EL, Richert-Boe KE, Henner WD.
Glutathione S-transferase M1 (GSTM l) deficiency and lung cancer risk. Cancer
Epidemiol Biomarkers Prev 1995;4z589-94.

Nazar-Stewart V, Vaughan TL, Burt RD, Chen C, Berwick M, Swanson GM.
Glutathione S-transferase M1 and susceptibility to nasopharyngeal carcinoma.
Cancer Epidemiol Biomarkers Prev 1999;82547-51.

London SJ DA, Cooper J et al. Polymorphism of glutathione S-transferase M1
and lung cancer risk among African-Americans and Caucasians in Los Angeles
county, California. J Natl Cancer Institute 1995;87:1246-1253.

Pemble S, Schroeder KR, Spencer SR, et al. Human glutathione S-transferase
theta (GS'I'TI): cDNA cloning and the characterization of a genetic
polymorphism. Biochem J 1994;300:271-6.

Hayes JD, Pulford DJ. The glutathione S-transferase supergene family: regulation
of GST and the contribution of the isoenzymes to cancer chemoprotection and
drug resistance. Crit Rev Biochem Mol Biol 1995;30:445-600.

Harries LW, Stubbins MJ, Forman D, Howard GC, Wolf CR. Identification of
genetic polymorphisms at the glutathione S- transferase Pi locus and association

with susceptibility to bladder, testicular and prostate cancer. Carcinogenesis
1997;18:641-4.

Pajarinen J, Savolainen V, Perola M, Penttila A, Karhunen PJ. Glutathione S-
transferase-Ml 'null' genotype and alcohol-induced disorders of human

spermatogenesis. Int J Androl 1996; 19: 155-63.

Ronis MJJ LKaI-SM. The CYP2E subfamily. In: DV ICaP, ed. Cytochromes
P450. Metabolic and toxicologic aspects. Boca Raton: CRC Press, 1996:211-239.

142

313.

314.

315.

316.

317.

318.

319.

320.

321.

322.

Baranova H, Canis M, Ivaschenko T, et al. Possible involvement of arylamine N-
acetyltransferase 2, glutathione S- transferases M1 and T1 genes in the
development of endometriosis. Mol Hum Reprod 1999;5z636-41.

Osteen KG, Sierra-Rivera E. Does disruption of immune and endocrine systems
by environmental toxins contribute to development of endometriosis? Semin
Reprod Endocrinol 1997;15:301-8.

Aitken RJ, Clarkson J S. Cellular basis of defective sperm function and its
association with the genesis of reactive oxygen species by human spermatozoa. J
Reprod Fertil 1987;81 :459-69.

Sharma RK, Agarwal A. Role of reactive oxygen species in male infertility.
Urology 1996;48:835-50.

Aitken RJ. Pathophysiology of human spermatozoa. Curr Opin Obstet Gynecol
1994;62128-35.

Twigg J, Fulton N, Gomez E, Irvine DS, Aitken RJ. Analysis of the impact of
intracellular reactive oxygen species generation on the structural and functional
integrity of human spermatozoa: lipid peroxidation, DNA fragmentation and
effectiveness of antioxidants. Hum Reprod 1998;13: 1429-36.

Twigg JP, Irvine DS, Aitken RJ. Oxidative damage to DNA in human
spermatozoa does not preclude pronucleus formation at intracytoplasmic sperm
injection. Hum Reprod 1998;13:1864-71.

de Lamirande E, Gagnon C. Capacitation-associated production of superoxide
anion by human spermatozoa. Free Radic Biol Med 1995;18:487-95.

Aitken RJ, Buckingham DW, West KM. Reactive oxygen species and human
spermatozoa: analysis of the cellular mechanisms involved in luminol- and
lucigenin-dependent chemiluminescence. J Cell Physiol 1992;151:466-77.

Voie OA, Wiik P, Fonnum F. Ortho-substituted polychlorinated biphenyls
activate respiratory burst measured as luminol-amplified chemoluminescence in
human granulocytes. Toxicol Appl Pharmacol 1998;150:369-75.

Ganey PE, Sirois JE, Denison M, Robinson JP, Roth RA. Neutrophil function
after exposure to polychlorinated biphenyls in vitro. Environ Health Perspect
1993;101:430-4.

Jacobson JL, Humphrey HE, Jacobson SW, Schantz SL, Mullin MD, Welch R.
Determinants of polychlorinated biphenyls (PCBs), polybrominated biphenyls

(PBBs), and dichlorodiphenyl trichloroethane (DDT) levels in the sera of young
children. Am] Public Health 1989;79:1401-4.

143

 

324.

325.

326.

327.

329.

330.

331.

332.

333.

335.

Falck F, J r., Ricci A, J r., Wolff MS, Godbold J, Deckers P. Pesticides and
polychlorinated biphenyl residues in human breast lipids and their relation to
breast cancer. Arch Environ Health 1992;47:143-6.

Kaminsky LS, Kennedy MW, Adams SM, Guengerich FP. Metabolism of
dichlorobiphenyls by highly purified isozymes of rat liver cytochrome P-450.
Biochemistry 1981;20:7379-84.

Kennedy MW CN, Dymerski PP et al. Metabolism of monochlorobiphenyls by
hepatic microsomal cytochrome P-450. Biochem Pharmacol 1980;29:727-736.

McLean MR, Robertson LW, Gupta RC. Detection of PCB adducts by the 32P-
postlabeling technique. Chem Res Toxicol 1996;9: 165-71.

Jones R, Mann T, Sherins R. Peroxidative breakdown of phospholipids in human
spermatozoa, spermicidal properties of fatty acid peroxides, and protective action
of seminal plasma. Fertil Steril 1979;31:531-7.

Nackerdien Z, Rao G, Cacciuttolo MA, Gajewski E, Dizdaroglu M. Chemical
nature of DNA-protein cross-links produced in mammalian chromatin by
hydrogen peroxide in the presence of iron or copper ions. Biochemistry
1991;30:4873-9.

Martins EA, Meneghini R. Cellular DNA damage by hydrogen peroxide is
attenuated by hypotonicity. Biochem J 1994;299:137-40.

Oluski R, Nackerdien, Z, and Dizdaroglu, M. DNA protein cross linking between
thymine and tyrosine in chromatin of gamma irradiated and hydrogen peroxide
treated cultured human cells. Arch Biochem Biophys 1992;297: 139-143.

Lee IP, Dixon RL. Effects of mercury on spermatogenesis studied by velocity
sedimentation cell separation and serial mating. J Pharmacol Exp Ther
1975;194:171-81.

Eamst E, and Lauritsen, JG. Effects of organic and inorganic mercury on human
sperm motility. Pharmacol Toxicol 1990;69:440-444.

Stacey NH, Kappus H. Cellular toxicity and lipid peroxidation in response to
mercury. Toxicol Appl Pharmacol 1982;63:29-35.

Alexidis AN, Rekka EA, Kourounakis PN. Influence of mercury and cadmium
intoxication on hepatic microsomal CYP2E and CYP3A subfamilies. Res
Commun Mol Pathol Pharmacol 1994;85:67-72.

144

336.

337.

338.

339.

342.

344.

345.

346.

347.

348.

Dickman MD, Leung KM. Mercury and organochlorine exposure from fish
consumption in Hong Kong. Chemosphere 1998;37:991-1015.

Sonzogni W. Maack L GT, Degenhardt D, Anderson H, and Fiore B.
Polychlorinated biphenyl congeners in blood of Wisconsin sport fish consumers.
Arch. Environ. Contam. Toxicol 1991;20:56-60.

Bimbaum LS. The role of structure in the disposition of halogenated aromatic
xenobiotics. Environ Health Perspect 1985;61:l 1-20.

Kafafi SA, Said HK, Mahmoud MI, Afeefy HY. The electronic and
thermodynamic aspects of Ah receptor binding. A new structure-activity model: I.
The polychlorinated dibenzo-p-dioxins. Carcinogenesis 1992;13: 1599-605.

Kafafi SA, Afeefy HY, Said HK, Kafafi AG. Relationship between aryl
hydrocarbon receptor binding, induction of aryl hydrocarbon hydroxylase and 7-
ethoxyresorufin O-deethylase enzymes, and toxic activities of aromatic
xenobiotics in animals. A new model. Chem Res Toxicol 1993;62328-34.

Canada N HW. Toxic chemicals in the Great Lakes and associated effects.
Environment Canada, Department of Fisheries and Oceans, Ottawa l991;2.

Sofikitis N, Miyagawa I, Dimitriadis D, Zavos P, Sikka S, Hellstrom W. Effects
of smoking on testicular function, semen quality and sperm fertilizing capacity. J
Urol 1995;154:1030-4.

Muller CH. Rationale, interpretation, validation, and uses of sperm function tests.
J Androl 2000;21:10-30.

Lipshultz LI, Caminos-Torres R, Greenspan CS, Snyder PJ. Testicular function
after orchiopexy for unilaterally undescended testis. N Engl J Med 1976;295:15-8.

Lipshultz LI. Cryptorchidism in the subfertile male. Fertil Steril 1976;27:609-20.

Alpert PF, Klein RS. Spermatogenesis in the unilateral cryptorchid testis after
orchiopexy. J Urol 1983;129:301-2.

Close CE, Roberts PL, Berger RE. Cigarettes, alcohol and marijuana are related to
pyospermia in infertile men. J Urol 1990;144:900-3.

Stillman RJ, Rosenberg MJ, Sachs BP. Smoking and reproduction. Fertil Steril
1986;46:545-66.

Vine MF, Margolin BH, Morrison HI, Hulka BS. Cigarette smoking and sperm
density: a meta-analysis. Fertil Steril 1994;61:35-43.

145

350. Van Thiel DH, Lester R, Sherins RJ. Hypogonadism in alcoholic liver disease:
evidence for a double defect. Gastroenterology 1974;67:1188-99.

351. Wilcox A, Weinberg C, Baird D. Caffeinated beverages and decreased fertility.
Lancet 1988;221453-6.

352. Anawalt BD, Bebb RA, Matsumoto AM, et a1. Serum inhibin B levels reflect
Sertoli cell function in normal men and men with testicular dysfunction. J Clin
Endocrinol Metab 1996;81:3341-5.

353. Jensen TK, Andersson AM, Hjollund NH, et al. Inhibin B as a serum marker of
spermatogenesis: correlation to differences in sperm concentration and follicle-

stimulating hormone levels. A study of 349 Danish men. J Clin Endocrinol Metab
1997;82:4059-63.

354. Klingmuller D, Haidl G. Inhibin B in men with normal and disturbed
spermatogenesis. Hum Reprod 1997;12:2376-8.

355. Pierik FH, Vreeburg JT, Stijnen T, De Jong FH, Weber RF. Serum inhibin B as a
marker of spermatogenesis. J Clin Endocrinol Metab 1998;83:3110-4.

356. Illingworth PJ, Groome NP, Byrd W, et a1. Inhibin-B: a likely candidate for the
physiologically important form of inhibin in men [see comments]. J Clin
Endocrinol Metab 1996;8 1 : 1321-5.

357. De Kretser DM. Morphology and physiology of the testis. In: KL B, ed.
Principles and practice of endocrinology and metabolism. Philadelphia: J B
Lippincott, 1995:1032 - 1041.

358. Pollard I. The testis and control of spermataogenesis. A guide to reproduction.
Cambridge: Cambridge University Press, 1994:74-93.

359. Winters SJ. Diurnal rhythm of testosterone and luteinizing hormone in
hypogonadal men. J Androl 1991;12:185-90.

360. Bremner WJ, Vitiello MV, Prinz PN. Loss of circadian rhythmicity in blood
testosterone levels with aging in normal men. J Clin Endocrinol Metab
1983;56:1278-81.

361. Krieger DT, Ossowski R, Fogel M, Allen W. Lack of circaidan periodicity of
human serum FSH and LH levels. J Clin Endocrinol Metab 1972;35:619-23.

362. Veldhuis JD, Iranmanesh A, Johnson ML, Lizarralde G. Twenty-four-hour
rhythms in plasma concentrations of adenohypophyseal hormones are generated
by distinct amplitude and/or frequency modulation of underlying pituitary
secretory bursts. J Clin Endocrinol Metab 1990;71:1616-23.

146

363.

364.

366.

370.

371.

372.

373.

374.

Barratt CL, Naeeni M, Clements S, Cooke ID. Clinical value of sperm
morphology for in-vivo fertility: comparison between World Health Organization
criteria of 1987 and 1992. Hum Reprod 1995;10:587-93.

Bostofte E, Bagger P, Michael A, Stakemann G. Fertility prognosis for infertile
men: results of follow-up study of semen analysis in infertile men from two
different populations evaluated by the Cox regression model. Fertil Steril
1990;54:1 100-6.

Barratt CL, St John JC. Diagnostic tools in male infertility. Hum Reprod 1998;13
Suppl 1:51-61.

Overstreet JW. Clinical approach to male reproductive problems. Occup Med
1994;92387-404.

Zanaveld LJD JR. Modern assessment of semen for diagnostic purposes. Semin
Reprod Endrocrinol 1988;62324-337.

Tummon IS, Yuzpe AA, Daniel SA, Deutsch A. Total acrosin activity correlates
with fertility potential after fertilization in vitro. Fertil Steril 1991;56:933-8.

Phillips DL, Pirkle JL, Burse VW, Bemert JT, Jr., Henderson LO, Needham LL.
Chlorinated hydrocarbon levels in human serum: effects of fasting and feeding.
Arch Environ Contam Toxicol 1989;18:495-500.

Patterson DG, J r., Needham LL, Pirkle JL, et a1. Correlation between serum and

adipose tissue levels of 2,3,7,8- tetrachlorodibenzo-p-dioxin in 50 persons from
Missouri. Arch Environ Contam Toxicol 1988;17:139-43.

Aarts J. Application of the chemical activated luciferase expression (CALUX)
bioassay for quantification of dioxin-like compounds in small samples of human
milk and bloodplasma. Organohalogen Compounds 1996;27:285-289.

Lancranjan I, Popescu, HI, Klepsch, I et al. Reproductive ability of workmen
occupationally exposed to lead. Arch Environ Health 1975;30:396-401.

Thomas J A, Brogan WCd. Some actions of lead on the sperm and on the male
reproductive system. Am J Ind Med 1983;42127-34.

Stewart P, Darvill T, Lonky E, Reihman J, Pagano J, Bush B. Assessment of
prenatal exposure to PCBS from maternal consumption of Great Lakes fish: an

analysis of PCB pattern and concentration. Environmental Research 1999;80:887-
S96.

147

375. Mullard AK, W and Wirth, JJ. Human exposure to polychlorinated biphenyls in
Michigan sport caught fish. 2001 (submitted).

376. Olsen J, Juul S, Basso 0. Measuring time to pregnancy. Methodological issues to
consider. Hum Reprod 1998;13:1751-3.

148

 

APPENDIX A

WHO GUIDELINES FOR SEMEN ANALYSIS PARAMETERS

149

Reference values
Volume

pH

Sperm concentration
Total sperm number
Motility

Morphology
Vitality

White blood cells
[mmunobead test

MAR test

2.0 ml or more
7.2 or more
20 X 106 spermatozoa/ml or more

. 40 X 10" spermatozoa per ejaculate or more

50% or more motile (grades a + b) or 2 5% or
more with . progressive motility (grade a)
within 60 minutes of ejaculation

75% or more live. i.e.. excluding dye

Fewer than 1 X 10"lml

Fewer than 50% motile spermatozoa with
beads bound

Fewer than 50% motile spermatozoa with
adherent particles

' Multicentre p0pulation-based studies utilizing the methods of mor-
phology assessment in this manual are now in~ progress.

Data from assisted reproductive technology programmes suggest
that. as sperm morphology falls below 1 5% normal forms using the
methods and definitions described in this manual. the fertilization rate

in vitro decreases.

150

APPENDIX B

SUMMARY OF CHEMICALS QUANTIFIED IN FISH TISSUE

TRIGGER LEVELS USED BY MDCH TO ESTABLISH CONSUMPTION
WARNINGS

151

688.38 «33 o... .0 ace 6 35.5.... «o: 893 $6590...
vcw 65.33.2588... 6553.0208qu 655332323 doc .39.... .onmé 33:... .2583: .:...v.<

3.08.. on .0: >2: 3.2, .2...a> 366.66 u

E3...» .30. 05 «a 35226:: «885:. u v.

 

 

66...... 9.3.65... 9.3 ...... - o... t ... cm... 66.:
6856 9.3626... 9.3 0." .. - ...3... v. o a 06698.. 66......
609.5 9.3696... 9.3 .....u - m8... .. an R 8.. .92.

“6.. 56565.62: 9.... o8... - .8... v. a 6 mm:

98.62: 96w 3. NS... . .8... v. n. .N 89.622698

9.... 6.62.9.3 68 I... - 8... on on 686.2

60.62: 9... .5... . .8... x ... R 6239268.:

6.9.55 9.3696... 9.3 8o... - 8c... 9. ... R 968m 6.69%:

66...; 9.3626... 9.3 8.... - m8... . ... a 6.6.:

9899.: 8.: 89 . m8... 9. E ..u .8:

893m 9.3696... 9.3 3.“ - «8... v. R R moo

98.6.: 9.... ..mv - 893. «a R on:

6856 9.3.696... 9.3 t6... - 8o... 9. on R 0690.666

.893... 33.696... 9.3 8e... - 8o... v. R R 56662.39.

6093.... 9.3696... 9.3 8m... - 83. x ... .8 6.6.62.6

6356 0.3.696... 9.3 3 ..o . 80... v. o. R 690.5...

c8936 9.3.626... 9.3 3.3 . 8.... v. .u R 866.6...
cos—5:00:00 62:.va .83. 3:3. 85:30 8.8.5.2

5.; 3.8% ...... 6.83 6.8.688 86. ..o a «9... .o e .8696

 

.3383 2.3.. gm: 55.8.0360 :. voczcwaw ...-8.82.0 .o 32:65

152

600030808 69.9.0.6 9.6. 0005.03 .98 0. 8...... ....6... .0. .96. .000... .082... :00... 9: ...

 

800 0.» 05:088..
800 0.0 3.0080300 02
500 9w 30> .0... 0.00.2 0
Eu: Nd 5:02 .0n. .00.: w
...... 00.0 .83 6n. .9... .

98> 0. 60.5 6.0.0.0
06 00.. 00.600 0.20 .0 60.02.

En... 0N 5.00.003. .8050
won .80..
. $00 0.0 .852
600 m; 30080030 02
800 ad 5:083:00 .0300”.
.6080...
E00 0.0 Aouuoqw .0.:08001L.020800I
.90.... u. .... 0.0 996.9600 9.6.. 66.0
. . 0.0.. 0.0 5.0.0.0
E... 0.0 ..00 .93
.93... u. E... 0.0 060.020 .90»

 

.96.. 600.... :00: 8.6.0

00.30.50 5:083:00 :0... 5.3800 8 5.00: .£55800 .0 325.0000 50.5.3 05 >0 000.. 0.0.6. .000...»

153

Mm“:

 

3