Exposure to persistent organochlorine pollutants associates
with human sperm Y:X chromosome ratio
Tarmo Tiido1,2,4, Anna Rignell-Hydbom3, Bo Jo ¨nsson3, Yvonne Lundberg Giwercman2,
Lars Rylander3, Lars Hagmar3and Aleksander Giwercman1
1Fertility Centre, Scanian Andrology Centre, and2Department of Urology, Malmo ¨ University Hospital, Lund University, SE 205 02
Malmo ¨ and3Department of Occupational and Environmental Medicine, Lund University Hospital, SE 21885, Lund, Sweden
4To whom correspondence should be addressed: Wallenberg Laboratory, entrance 46, Malmo ¨ University Hospital, SE 205 02 Malmo ¨,
Sweden. E-mail: Tarmo.Tiido@kir.mas.lu.se
BACKGROUND: During the last decades, there has been concern that exposure to endocrine disruptors, such as
persistent organochlorine pollutants (POPs), may contribute to sex ratio changes in offspring of exposed popu-
lations. METHODS: To investigate whether exposure to 2,204,405,50-hexachlorobiphenyl (CB-153) and dichlorodi-
phenyl dichloroethene (p,p0-DDE) affect Y:X chromosome proportion, semen of 149 Swedish fishermen, aged 27–
67 years, was investigated. The men provided semen and blood for analysis of hormone, CB-153 and p,p0-DDE
levels. The proportion of Y- and X-chromosome bearing sperm in semen samples was determined by two-colour
fluorescence in situ hybridization (FISH) analysis. RESULTS: Log transformed CB-153 as well as log transformed
p,p0-DDE variables were both significantly positively associated with Y chromosome fractions (P-values 5 0.05 and
<0.001, respectively). Neither age, smoking nor hormone levels showed any association with Y-chromosome frac-
tions. CONCLUSIONS: This is the first study to indicate that exposure to POPs may increase the proportion of
ejaculated Y-bearing spermatozoa. These data add to the growing body of evidence that exposure to POPs may
alter the offspring sex ratio.
Key words: polychlorinated biphenyls/POP/p,p0-DDE/sex ratio/sperm
Worldwide, the human sex ratio at birth is fairly constant—
the male proportion of all births being 51.4% (James, 1996a).
However, recent studies indicate that in many countries the
proportion of male births has been declining during the past
five decades (Allan et al., 1997; van der Pal-de Bruin et al.,
1997; Marcus et al., 1998; Møller, 1998; Parazzini et al.,
1998). In order to explain this phenomenon several hypoth-
eses have been put forward, including influence of parental
hormone levels (James, 1996a, 2001), changes in coital rates
(Dickinson and Parker, 1996), the length of inter-pregnancy
intervals (James, 1996b), or changes in maternal nutrition
(Rosenfeld et al., 2003). It has also been suggested that the
time-related reduction in male proportion could be due to an
increasing exposure to ‘endocrine disrupters’ such as persist-
ent organohalogen pollutants (POPs) (Toppari et al., 1996).
POPs, such as polychlorinated dibenzofurans (PCDFs),
polychlorinated dibenzo-p-dioxins (PCDD), polychlorinated
biphenyls (PCBs), dichlorodiphenyl trichloroethane (DDT),
and dichlorodiphenyl dichloroethene (p,p0-DDE, the most
stable daughter compound of DDT), are ubiquitous environ-
mental contaminants. These compounds are highly persistent,
which results in bioaccumulation and biomagnification in the
food chain. Studies have shown that measurable levels of
PCBs and p,p0-DDE are found in a large proportion of the
general population (Longnecker et al., 1997). The half-lives
of different PCB congeners in the blood range from 1 to 10 or
more years, while p,p0-DDE has a half-life of $10 years
(Brown et al., 2003). Some of these chemicals can disrupt
multiple endocrine pathways and induce a wide range of toxic
responses (Toppari et al., 1996). A variety of studies have
demonstrated their estrogenic, anti-estrogenic, and androgen
competing properties (Kelce et al., 1995; Danzo, 1997). Two
accidents, which have attracted a lot of attention, are the
Yucheng poisoning (Chen et al., 1985; Masuda et al., 1985)
and the Seveso disaster (Mocarelli et al., 1996), both of
which were associated with an increased proportion of girls
born subsequently to paternal exposure to POPs (Mocarelli
et al., 1996, 2000; del Rio Gomez et al., 2002). In human
populations exposed to more moderate levels of POPs, both
increased (Karmaus et al., 2002) and decreased (Rylander
et al., 1995) male: female sex ratios have been reported.
Therefore, the explanation of the secular trend in sex ratio is
still lacking and the mechanisms that can affect the proportion
of males to females are not yet understood.
Theoretically, offspring sex ratio determination may be
attributed to events that occur before fertilization that favor
selection of Y- or X-chromosome bearing spermatozoa,
Human Reproduction Vol.20, No.7 pp.1903–1909, 2005
Advance Access publication April 28, 2005
q The Author 2005. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: firstname.lastname@example.org
events that occur after fertilization such as preferential devel-
opment or survival of embryos of one sex, or a combination
of both. In case the paternal exposure is crucial, as supported
by all studies to date (Mocarelli et al., 1996; Mocarelli et al.,
2000; del Rio Gomez et al., 2002), changes in male germ
cell development or function could be causative of the sex
ratio changes observed. Although recent human studies have
indicated that paternal exposure to POPs has a deleterious
effect on sperm parameters (Guo et al., 2000; Hauser et al.,
2003; Richthoff et al., 2003) it is not yet known whether
these compounds could change the proportion of X- and
In Sweden the main exposure route for POP is through
consumption of contaminated fatty fish from the Baltic Sea
of the eastern coast of Sweden (Svensson et al., 1995;
Rylander and Hagmar, 1999). Swedish fishermen constitute a
socio-economically homogeneous group with high fish con-
sumption and previously, east as well as west coast fishermen
have been found to eat on average more than twice as much
locally caught fatty fish than subjects from the general
Swedish population. This has resulted in increased POP
levels in plasma among east coast fishermen compared to
west coast fishermen as well as the general population
(Svensson et al., 1995). In the current study, east coast fisher-
men constituted ‘the more exposed group’ and west coast
fishermen ‘the less exposed group’. The choice of study base
ensured us of sufficient variation in POP exposure.
Reliable biomarkers of POP exposure are necessary to
establish dose–response relationships. The PCB congener,
2,20,4,405,50-hexachlorobiphenyl (CB-153), found in rela-
tively high concentrations in human serum, was selected as a
biomarker for POP exposure due to its very high correlations
with the total PCB concentration (Grimvall et al., 1997;
Glynn et al., 2000), the 2,3,7,8-tetrachlorodibenzo-p-dioxin
(TCDD) equivalent (TEQ) from PCB, and the total POP
derived TEQ (Gladen et al., 1999), respectively. Another rel-
evant biomarker is the anti-androgenic compound p,p0-DDE,
which is present in relatively high serum concentrations in
men consuming fatty fish from the Baltic Sea (Sjo ¨din et al.,
In order to enlighten the mechanism behind possible link
between POP exposure and offspring sex ratio, we aimed to
investigate whether there is an association between serum
levels of CB-153 and p,p0-DDE and the Y:X chromosome
ratio in ejaculated sperms of Swedish fishermen.
Materials and methods
Cohorts of fishermen from the Swedish east and west coasts were
established in 1988 (Svensson et al., 1995). In year 2000, a postal
questionnaire, mainly focused on fracture incidence, was sent to
3505 west coast fishermen and 1678 east coast fishermen from the
original cohorts, born 1935 or later. The questionnaire included a
question about whether the subjects were interested in more infor-
mation on a study of male semen function. Among the 2614 subjects
(east n ¼ 848 and west n ¼ 1766) who responded to this specific
question, 479 (east n ¼ 171 and west n ¼ 308) wanted more
information about the semen study. We contacted these subjects as
well as another 169 east coast fishermen that had become members
of the east coast fishermen’s organization after the closure of the
cohorts, with written detailed information. From the east coast,
130 out of 340 men wanted to participate, and from the west coast
136 out of 308. During the field study period, 71 subjects were
excluded due to logistical reasons, changes of mind, sickness or
The non-participants from the original fishermen’s cohort had
similar age distribution as the participants (Rignell-Hydbom et al.,
Swedish Medical Birth Register supports that there was no differ-
ence in number of fathered children between participants and
Out of 195 men who participated in the semen study, 183 donated
enough semen for FISH analyses. Out of these 183, 28 samples
were excluded due to low number of cells available or failure during
analysis. No statistically significant differences regarding age, lipid-
adjusted levels of CB-153 and p,p0-DDE, percentage of A þ B
motile sperms, sperm concentration or total sperm count were found
between the 155 participating men with enough semen and the 28
subjects who were excluded due to failure of FISH (21 subjects
with hybridization failure and seven with low sperm number).
Exposure data were lacking in six men; the results are thus based on
based ondata fromthe
Mobile laboratory unit, semen and blood sampling, and
A mobile laboratory unit was established for collection and analysis
of semen and blood samples (Rignell-Hydbom et al., 2004). The
subjects were informed to keep an abstinence period of 3–4 days
before collection and in each case the actual length of abstinence
period was recorded (median 3.0, range 0.5–15). Sperm motility
and sperm concentration were analysed within 1h after collection
according to World Health Organization guidelines (WHO, 1999).
Undiluted raw semen was transferred into two tubes and directly put
into a box with dry ice. The semen samples were, thereafter, stored
in a freezer at 2808C until analysis. Venous blood samples were
collected between 7 am and 9 pm. The samples were centrifuged
and sera were frozen at 2808C for later analysis.
Information on lifestyle and reproductive history (e.g. number and
sex of children) was collected through telephone interviews. The
questionnaire was sent to the participants a couple of weeks before
telephone contact. During the telephone contact, an agreement was
reached on time and date for collection of semen and blood samples
at the subject’s home. The participants received verbal and written
information on the procedures for collecting the semen samples.
The fraction of current smokers was 21% (n ¼ 31), and of former
smokers 42% (n ¼ 62) among the participants. Other background
characteristics of the study population are given in Table I. Sperm
characteristics of this population have previously been reported
(Rignell-Hydbom et al., 2004). No men presented with azoospermia.
(median: 50.1 £ 106/ml).
5.7and207 £ 106/ml
Preparation of spermatozoon nuclei
After thawing and mixing, 10ml of semen was smeared on cleaned
microscope slides (Superfrost Plus slides, Menzel Gla ¨ser, Germany).
The smear was air-dried at room temperature for 24h. Sperm nuclear
decondensation was performed by incubation in 10mM dithiothreitol
(DTT; Saveen Werner AB, Sweden)/0.1M Tris for 7min, followed
by incubation in 1mM DDT/4mM lithium diiodosalicylate (Sigma-
Aldrich Chemie GmbH, Steinheim, Germany)/0.1M Tris for 20min.
T.Tiido et al.
Slides were then washed in 2 £ saline sodium citrate (SSC) and
Two-colour fluorescence in situ hybridization (FISH)
Prior to hybridization, the slides were incubated for 1min in a pre-
treatment solution containing 85mg/ml pepsin (Sigma-Aldrich Che-
mie GmbH, Steinheim, Germany) in 0.01M HCl, rinsed in TN
buffer (0.1M Tris–HCl, 0.15M NaCl) for 15min, and air-dried. To
determine the proportion of X and Y chromosomes, two-colour
FISH was performed using protein–nucleic acid (PNA) probes (pro-
vided by DakoCytomation, Glostrup, Denmark) targeted against the
centromeric region of the X chromosome (Rhodamine-labelled) and
the q-arm of the Y chromosome (FITC-labelled). The probe mixture
was placed on the semen smears within a marked area of a slide,
mounted with a cleaned cover slip and sealed with rubber cement.
Subsequently, probe- and target DNA were denatured simul-
taneously at 738C for 2min. Following overnight hybridization at
378C in a humid chamber, slides were washed twice, 5min each, at
368C in 60% formamide/2 £ SSC, in 0.2 £ SSC for 5min at room
temperature, and 5min in TN buffer/0.05% Tween 20. Slides were
rinsed in 2 £ SSC, and dehydrated in ethanol series (70–
90–100%). Thereafter,the slides
0.1mg/ml of 40,6-diamidino-2-phenylindole (DAPI) for 30s, and
dehydrated. Finally, the slides were mounted in Vectashield antifade
medium (Vector Laboratories Inc., Burlingame, USA).
Microscopy and scoring criteria
Slides were examined with an Olympus AX 70 epifluorescence
microscope (magnification: 400£) equipped with a single and
double band pass filter to detect DAPI, FITC and Rhodamine. The
examination was performed blindly, i.e. without knowledge of the
exposure levels or other subject characteristics. An X or Y chromo-
some in a sperm nucleus was recognized by a red or a green fluor-
escent spot, respectively. Sperm nuclei were only scored when
morphologically preserved, not clumping or overlapping, showing
well defined outline and tail and sperm head decondensed to no
more than twice the size of normal non-decondensed spermatozoa.
Sperm showing two signals (disomic or diploidic) were not counted.
Sperm cells showing normal decondensation, but without signals,
were scored and considered valid for calculation of hybridization
efficiency. In every sample the proportion of sperms presenting with
a clear red or green signal was $95%. There was no correlation
between the proportion of unlabelled spermatozoa and the fraction
of Y-chromosome bearing sperm (r ¼ 20.005, P ¼ 0.9).
The Y/X chromosome status of the spermatozoa was evaluated by
assessing randomly selected visual fields. The number of cells
which were evaluated ranged between 276 and 1301. This variation
was due the fact that a quality control program was included in the
first part of the study, in order to assess the inter- and intra-observer
coefficient of variation (CV). In several previous studies addressing
the issue of sperm aneuploidy, 10000 sperm per sample were
counted. However, since we were looking for more common event
than aneuploidy, our a priori assumption was that an acceptable
inter- and intra-observer CV could be achieved with scoring of a
lower number of spermatozoa. In the first part of the study, two
investigators assessed inter-observer variation by counting ,500
and ,1000 spermatozoa in each of 29 slides. Thereafter, one of the
investigators (T.T.) assessed intra-observer variation in 10 slides
twice by counting 500 and 1000 cells. Proportions of Y-chromo-
some bearing spermatozoa from these assessments are based on
counts of 1000 cells (median 1042; range 881–1301). Thereafter,
inter- and intra-observer CV as considers the proportion of
Y-chromosome bearing sperm, was estimated to be 2.3% and 3.3%,
respectively, by scoring 500 cells only and this procedure was sub-
sequently applied. However, due to the quality of the samples in six
cases the number of nuclei scored was below 500. The proportion of
spermatozoa with Y chromosome did not differ between these
samples and the remaining 149.
Serum concentrations of follicle stimulating hormone (FSH), lutei-
nizing hormone (LH) and estradiol were analysed with immuno-
fluorometric techniques. The total assay CVs were 2.9%, 2.6% and
8.1%, respectively. Serum testosterone and sex hormone-binding
globulin (SHBG) were measured by commercially available immu-
noassays. The total assay CVs were 5.5% and 4.6%, respectively.
Inhibin B levels were assessed using a specific immunometric
method, as previously described (Groome et al., 1996), with a detec-
tion limit of 15ng/l and intra-assay and total assay CVs ,7%.
Determination of CB-153 and p,p0-DDE
The levels of CB-153 were determined as previously described
(Richthoff et al., 2003). In addition, p,p0-DDE was analysed by the
same method. Briefly, the CB-153 and p,p0-DDE were extracted
from the serum by solid phase extraction (Isolute ENVþ; IST, Hen-
goed, UK) using on-column degradation of the lipids and analysis
by gas chromatography mass spectrometry.
and13C-labelled p,p0-DDE were used as an internal standards. The
selected ion monitoring of p,p0-DDE was performed at m/z 318
while m/z 330 was used for the internal standard. The relative stan-
dard deviations, calculated from samples analysed in duplicate at
different days, was 7% at 0.6ng/ml (n ¼ 76) and 5% at 1.5ng/ml
(n ¼ 37) for CB-153 and 12% at 0.6ng/ml (n ¼ 56) and 7% at
2.4ng/ml (n ¼ 50) for p,p0-DDE. The detection limits were
0.05ng/ml for CB-153 and 0.1ng/ml for p,p0-DDE. The analyses of
CB-153 and p,p0-DDE are part of the Round Robin inter-comparison
program (Hans Drexler, Institute and Out-Patient Clinic for
Occupational, Social and Environmental Medicine, University of
Erlangen-Nuremberg, Germany) with analysis results within the
Determination of lipids by enzymatic methods
Serum concentrations of triglycerides, cholesterol and phospholipids
were determined by enzymatic methods using reagents from
Boehringer–Mannheim (triglycerides and cholesterol; Mannheim,
Germany) and Waco Chemicals (phospholipids; Neuss, Germany).
Table I. Outcome, exposure and potential cofounders of the study
population of Swedish fishermen (n ¼ 149)
Fraction of Y
CB-153 (ng/g lipids)
p,p0-DDE (ng/g lipids)
Abstinence time (days)
POPs and sperm sex chromosome ratio
The total lipid concentration in serum was calculated by summation
of the amounts of triglycerides, cholesterol and phospholipids. In
these calculations, the average molecular weights of triglycerides
and phospholipids were assumed to be 807 and 714. For cholesterol,
we used an average molecular weight of 571, assuming that the
proportion of free and esterified cholesterol in serum was 1:2.
In linear regression models, we evaluated the effect of the exposure
variables CB-153 and p,p0-DDE serum concentrations, respectively,
on the fraction of Y chromosomes. The CB-153 and p,p0-DDE vari-
ables were analysed as continuous (untransformed and log trans-
formed) as well as categorized variables (into five equally sized
groups). Due to the very high correlation between CB-153 and p,p0-
DDE (r ¼ 0.73) in the present dataset, both variables were not taken
into the models simultaneously. Age (as continuous), current smok-
ing (yes/no), abstinence time (0–2, .2–4, .4–6, and .6 days)
and paternal sex hormone levels (testosterone, FSH, LH, SHBG,
and testosterone/SHBG ratio, all as continuous variables), were
suggested to affect offspring sex ratio (Hilsenrath et al., 1997;
James, 2001; Fukuda et al., 2002), and were therefore considered as
potential confounders. If these variables showed any association
(P , 0.20) with the Y chromosome fraction they were included in
the models, one at a time, together with the exposure variable. If the
adjusted estimates differed by ,15% from the crude estimate, only
crude results are presented. The model assumption was checked by
means of residual analysis. In addition, to ensure that linear associ-
ations were reasonable, scatter plots were evaluated for all bivariate
The log transformed lipid adjusted p,p0-DDE concentration
was significantly (P , 0.001) associated with the Y chromo-
some fraction (slope [b] for ln[p,p0-DDE] 0.66, 95% confi-
dence interval [CI] 0.30, 1.01; Figure 1). According to the
regression model this means that a p,p0-DDE concentration
of 242ng/g lipid (the median level) corresponds to a Y
chromosome fraction of 51.2% and a p,p0-DDE concentration
of 472ng/g lipid (lower limit value for highest exposure cat-
egory) corresponds to 51.7%. However, p,p0-DDE explains
only 7.5% of the total variance for Y:X chromosome
The log transformed lipid adjusted CB-153 concentration
was also significantly (P ¼ 0.05) associated with the Y
chromosome fraction (b for ln[CB-153] 0.42, 95% confi-
dence interval [CI] 0.01, 0.83; Figure 2). According to the
regression model this means that a CB-153 concentration of
200ng/g lipid (the median level) corresponds to a Y chromo-
some fraction of 51.3% and a CB-153 concentration of
328ng/g lipid (lower limit value for highest exposure
category) corresponds to 51.5%. However, CB-153 explains
only 2.0% of the total variance for Y:X chromosome
For CB-153, and especially for p,p0-DDE, the log trans-
formed variables better fulfilled model assumptions as com-
pared with the untransformed ones. When the exposure
variables were divided into five categories, subjects in the
category with the lowest concentration of p,p0-DDE had a
significantly lower Y chromosome fraction as compared with
the category with the highest exposure (mean difference
1.6%, 95% CI 0.8, 2.5, P , 0.001; Figure 3). A correspond-
ing comparison for CB-153 showed a tendency of a lower
fraction of Y chromosomes among subjects within the lowest
exposure category as compared with the subjects in the cat-
egory with the highest exposure (mean difference 0.8%, 95%
CI–0.1, 1.7, P ¼ 0.07; Figure 4).
Neither smoking (P ¼ 0.24) nor hormone parameters
(all P-values .0.40) showed any association with the Y
Figure 2. The association between 2,20,4,40,5,50-hexachlorobiphenyl
(CB-153), and the fraction of sperm with Y chromosomes in 149
fishermen from Sweden. The regression model gave a predicted
fraction of Y chromosomes of 49.0 þ 0.42*ln(CB-153) (P ¼ 0.05).
Figure 1. The association between dichlorodiphenyl dichloroethene
(p,p0-DDE) and the fraction of sperm with Y chromosomes in
149 fishermen from Sweden. The regression model gave a
predicted fraction of Y chromosomes of 47.6 þ 0.66*ln(p,p0-DDE)
(P , 0.001).
T.Tiido et al.
chromosome fractions and the same was noted regarding age,
irrespective of whether age was considered as a continuous
(P ¼ 0.47) or categorical (P ¼ 0.30) variable with the cut-off
points at median age. Hence, none of these parameters were
included in the multivariate models. On the other hand, men
with the longest abstinence period tended to have a some-
what lower proportion of Y-chromosome bearing sperms as
compared with the rest. When including abstinence period in
the models, the effect estimates were, however, only changed
The Y chromosome fractions did not differ (P ¼ 0.36)
between men who had fathered only/more boys (n ¼ 59) as
compared with men who had fathered only/more girls
(n ¼ 37).
Our main finding was that a higher exposure to POPs,
expressed as concentrations of CB-153 and p,p0-DDE in
serum, was associated with a slightly higher proportion of Y-
chromosome bearing sperms. Due to the strong correlation
between the serum levels of the two POP biomarkers, we
can, however, not assess whether the observed effect was
linked to one or both of these exposure markers, or to any
other correlated POP.
When evaluating the results of the current study, several
potential biases need to be considered. It has been shown that
the participation rate in sperm studies is related to both age
and experienced fertility (Larsen et al., 1998). In the present
study, the age-distributions as well as the mean number of
children were very similar among the participants and the
non-participants (Rignell-Hydbom et al., 2004). Therefore,
we do not consider that selection bias is of major concern.
Moreover, we believe that residual confounding is probably
not an issue of great concern, as we have considered all
potential confounders known to us in the analyses. However,
we cannot exclude that imperfect measurements of the con-
founders have caused some residual confounding. Only hav-
ing a single semen sample could be a limitation of this study.
However, there are no data in the literature on possible
temporal variation in chromosomal fractions. Even if such a
phenomenon exists, this variability would most likely dilute
and not magnify the associations found in the present study.
Another potential bias, which needs to be considered, was
the lack of hybridization signal in up to 5% of spermatozoa.
This could be attributed to insufficient hybridization. How-
ever, a hybridization efficiency of 95% or more is in good
accordance with the hybridization efficiency reported by
other groups (e.g. Martin et al., 1996; Johannisson et al.,
2002). Since the proportion of unlabelled spermatozoa did
not correlate to the proportion of those with Y chromosome,
it seemed probable that cell-based hybridization failure
shared equally between the Y bearing and X bearing cells,
and hence no additional probe for any autosomal chromo-
some was applied. Moreover, the results of our study could
hardly be influenced by sex chromosomal aneuploidies, since
previous studies have shown that sex chromosomal aneuploi-
dies are rare, occurring in ,0.5% of cells (Martin et al.,
1996). Twenty-eight cases were excluded due to insufficient
labelling. However, there was no significant difference con-
sidering age, exposure level, and seminal parameters between
subjects from whom FISH data were obtained and those
excluded from the study.
To our knowledge, this is the first study to elucidate the
impact of POP exposure on sex chromosome ratio in sperm
and our results indicate that sperm Y:X ratio is influenced by
exposure to POPs. The design of our study does not, how-
ever, allow us to clarify the mechanisms behind the observed
association. One hypothesis could be that testicular apoptosis,
which was reported to be sex hormone regulated (Billig et al.,
Figure 3. The association between dichlorodiphenyl dichloroethene
(p,p0-DDE) and the fraction of sperm with Y chromosomes in 149
fishermen from Sweden. Mean values and 95% confidence intervals
Figure 4. The association between 2,20,4,40,5,50-hexachlorobiphenyl
(CB-153) and the fraction of sperm with Y chromosomes in 149
fishermen from Sweden. Mean values and 95% confidence intervals
POPs and sperm sex chromosome ratio
1996), might also be controlled through the effect of POPs
mediated via androgen or estrogen signalling, as POPs are
able to act as sex hormone receptor agonists or antagonists
(Kelce et al., 1995; Danzo, 1997). Furthermore, it was
recently shown that the aryl hydrocarbon receptor (AHR)
which mediates the toxic effect of dioxins and regulates
apoptosis of germ cells was highly expressed in human testis
(Schultz et al., 2003). However, it is not clear how or
whether apoptosis might specifically change the ratio
between Y- and X-chromosome bearing spermatids. Another
possible mechanism could be a loss of the X chromosome
through an organochlorine effect on formation of micronuclei
during the process of meiosis. Organochlorines were shown
to have an effect on micronucleation in somatic cells both
in vitro and in vivo (Cicchetti et al., 1999; Gauthier et al.,
1999; Lu et al., 2000), whereas a similar effect on meiotic
cells has not yet been shown.
It is a reasonable hypothesis, as yet unproven however,
that the Y:X chromosome sperm ratio will affect the male:
female offspring ratio. There are several epidemiological
studies evaluating the effect of different POPs on the sex
ratio outcome, giving a rather ambiguous picture. Accidental
high paternal TCDD exposure from the Seveso disaster sig-
nificantly lowered the sex ratio among the offspring several
years later (Mocarelli et al., 1996, 2000), but on the other
hand more long-term but lower exposure levels to dioxins
among US Vietnam veterans showed a tendency towards an
increased proportion of boys fathered (Michalek et al., 1998).
The Yucheng poisonings by PCB and PCDF lead to a
decrease in the proportion of boys born, but only among off-
spring to men exposed before adulthood (del Rio Gomez
et al., 2002), whereas on the other hand, there was among
Michigan fish eaters an association between paternal serum
PCB concentrations and increased odds ratio of fathering a
boy (Karmaus et al., 2002). On a cohort basis, Baltic Sea
fishermen’s families from the Swedish east coast had higher
POP levels in blood than fishermen’s families from the west
coast (Rylander and Hagmar 1995; Svensson et al., 1995).
The proportion of boys was significantly lower among the
offspring from the east coast cohort as compared to the west
coast cohort; none of these figures differed, however, signifi-
cantly from the sex ratio of the overall Swedish population
(Rylander et al. 1995). It is possible that the inconsistent
associations between POP exposure and sex ratio are due to
different impact on the male reproductive system of low-
level continuous and major, but short-term, exposures,
respectively. Furthermore, apart from the dose and the dur-
ation of exposure, differing sex hormone mimicking effects
of the various POPs, as shown in vitro, might contribute to
diverging in vivo effects (Bønefeld-Jorgensen et al., 2001).
The relative composition of POP compounds has varied con-
siderably between the different exposure situations where the
sex ratio has been assessed. Our findings on positive corre-
lation between the proportion of Y-bearing sperms and serum
levels of p,p0-DDE and CB-153 could be considered to be in
concordance with the work of Karmaus et al. (2002), linking
low-level continuous paternal POP exposure to offspring sex
In the current study, the Y chromosome fraction did not
differ between the 59 men who fathered only/more boys as
compared to those 37 men who fathered only/more girls.
However, it should be kept in mind that the number of chil-
dren of exposed fishermen in the study is small and even if
there were a direct link between sperm Y:X ratio and sex
ratio of the offspring, the statistical power would be insuffi-
cient for detecting any effect on sex ratio of the presently
observed weak association between POP exposure and sperm
Y:X ratio. Furthermore, the semen samples were collected
several years after the birth of the children and it is unclear
how stable Y:X ratios are over time.
In conclusion, our observed effects of a slightly higher
Y:X sperm chromosome ratio among men with higher POP
concentrations in serum do not give any circumstantial evi-
dence in support for the hypothesis that the trend of declining
sex ratio in several societies over past decades (Parazzini
et al., 1998) has been due to increasing exposure to POP.
However, the relation between paternal Y:X sperm chromo-
some ratio and offspring sex ratio may be quite complicated.
Even factors like timing of exposure in sensitive periods like
fetal life and puberty need to be taken into consideration.
Elucidation of the intratesticular mechanisms affecting the
distribution of Y and X sperm may not only add to our
understanding of the biological mechanisms regulating the
offspring sex ratio but also contribute to better understanding
of the process of spermatogenesis.
We are indebted to Camilla Anderberg for help in FISH analysis
and Kirsten Vang Nielsen from DakoCytomation for providing the
PNA probes. This work was supported by grants from the European
Commission (QLK4-CT-2001-00202), Swedish Research Council
(Grant No 521-2002-3907), the Swedish Research Council for
Environment, Agricultural Sciences and Spatial Planning, Swedish
government founding for clinical research and Crafoordska Fund.
Allan BB, Brant R, Seidel JE and Jarrell JF (1997) Declining sex ratios in
Canada. CMAJ 156,37–41.
Billig H, Chun SY, Eisenhauer K and Hsueh AJ (1996) Gonadal cell apopto-
sis: hormone-regulated cell demise. Hum Reprod Update 2,103–117.
Bønefeld-Jorgensen EC, Andersen HR, Rasmussen TH and Vinggaard AM
(2001) Effect of highly bioaccumulated polychlorinated biphenyl conge-
ners on estrogen and androgen receptor activity. Toxicology 158,141–153.
Brown RJ, Rundle SD, Hutchinson TH, Williams TD and Jones MB (2003)
A copepod life-cycle test and growth model for interpreting the effects of
lindane. Aquat Toxicol 63,1–11.
Chen PH, Wong CK, Rappe C and Nygren M (1985) Polychlorinated biphe-
nyls, dibenzofurans and quaterphenyls in toxic rice-bran oil and in the
blood and tissues of patients with PCB poisoning (Yu-Cheng) in Taiwan.
Environ Health Perspect 59,59–65.
Cicchetti R, Bari M and Argentin G (1999) Induction of micronuclei in bone
marrow by two pesticides and their differentiation with CREST staining:
an in vivo study in mice. Mutat Res 439,239–248.
Danzo BJ (1997) Environmental xenobiotics may disrupt normal endocrine
function by interfering with the binding of physiological ligands to steroid
receptors and binding proteins. Environ Health Perspect 105,294–301.
del Rio Gomez I, Marshall T, Tsai P, Shao YS and Guo YL (2002) Number
of boys born to men exposed to polychlorinated biphenyls. Lancet 360,
Dickinson HO and Parker L (1996) Why is the sex ratio falling in England
and Wales? J Epidemiol Community Health 50,227–228.
T.Tiido et al.
Fukuda M, Fukuda K, Shimizu T, Andersen CY and Byskov AG (2002) Par- Download full-text
ental periconceptional smoking and male: female ratio of newborn infants.
Gauthier JM, Dubeau H and Rassart E (1999) Induction of micronuclei in
vitro by organochlorine compounds in beluga whale skin fibroblasts.
Mutat Res 439,87–95.
Gladen BC, Longnecker MP and Schecter AJ (1999) Correlations among
polychlorinated biphenyls, dioxins, and furans in humans. Am J Ind
Glynn AW, Wolk A, Aune M, Atuma S, Zettermark S, Maehle-Schmid M,
Darnerud PO, Becker W, Vessby B and Adami HO (2000) Serum concen-
trations of organochlorines in men: a search for markers of exposure.
Sci Total Environ 263,197–208.
Grimvall E, Rylander L, Nilsson-Ehle P, Nilsson U, Stro ¨mberg U, Hagmar L
and O¨stman C (1997) Monitoring of polychlorinated biphenyls in human
blood plasma: methodological developments and influence of age, lacta-
tion, and fish consumption. Arch Environ Contam Toxicol 32,329–336.
Groome NP, Illingworth PJ, O’Brien M, Pai R, Rodger FE, Mather JP and
McNeilly AS (1996) Measurement of dimeric inhibin B throughout the
human menstrual cycle. J Clin Endocrinol Metab 81,1401–1405.
Guo YL, Hsu PC, Hsu CC and Lambert GH (2000) Semen quality after pre-
natal exposure to polychlorinated biphenyls and dibenzofurans. Lancet
Hauser R, Chen Z, Pothier L, Ryan L and Altshul L (2003) The relationship
between human semen parameters and environmental exposure to poly-
chlorinated biphenyls and p,p0-DDE. Environ Health Perspect 111,
Hilsenrath RE, Swarup M, Bischoff FZ, Buster JE and Carson SA (1997)
Effect of sexual abstinence on the proportion of X-bearing sperm as
assessed by multicolor fluorescent in situ hybridization. Fertil Steril 68,
James WH (1996a) Evidence that mammalian sex ratios at birth are partially
controlled by parental hormone levels at the time of conception. J Theor
James WH (1996b) Interpregnancy intervals, high maternal age and seasonal
effects on the human sex ratio. Hum Reprod 11,7–8.
James WH (2001) The data sources which may help strengthen the epide-
miological evidence for the hormonal hypothesis of sex determination in
man. Hum Reprod 16,1081–1085.
Johannisson R, Leuschner E, Huppe M, Hinrichs F, Al Hasani S, Diedrich K,
Schwinger E and Mennicke K (2002) Increased frequency of X-bearing
sperm in males from an infertility clinic: analysis by two-color fluor-
escence in situ hybridization. Cytogenet Genome Res 98,240–244.
Karmaus W, Huang S and Cameron L (2002) Parental concentration of
dichlorodiphenyl dichloroethene and polychlorinated biphenyls in Michi-
gan fish eaters and sex ratio in offspring. J Occup Environ Med 44,8–13.
Kelce WR, Stone CR, Laws SC, Gray LE, Kemppainen JA and Wilson EM
(1995) Persistent DDT metabolite p,p0-DDE is a potent androgen receptor
antagonist. Nature 375,581–585.
Larsen SB, Abell A and Bonde JP (1998) Selection bias in occupational
sperm studies. Am J Epidemiol 147,681–685.
Longnecker MP, Rogan WJ and Lucier G (1997) The human health effects
of DDT (dichlorodiphenyltrichloroethane) and PCBS (polychlorinated
biphenyls) and an overview of organochlorines in public health. Annu Rev
Public Health 18,211–244.
Lu Y, Morimoto K, Takeshita T, Takeuchi T and Saito T (2000) Genotoxic
effects of alpha-endosulfan and beta-endosulfan on human HepG2 cells.
Environ Health Perspect 108,559–561.
Marcus M, Kiely J, Xu F, McGeehin M, Jackson R and Sinks T (1998)
Changing sex ratio in the United States, 1969-1995. Fertil Steril 70,
Martin RH, Spriggs E and Rademaker AW (1996) Multicolor fluorescence in
situ hybridization analysis of aneuploidy and diploidy frequencies in
225,846 sperm from 10 normal men. Biol Reprod 54,394–398.
Masuda Y, Kuroki H, Haraguchi K and Nagayama J (1985) PCB and PCDF
congeners in the blood and tissues of yusho and yu-cheng patients.
Environ Health Perspect 59,53–58.
Michalek JE, Rahe AJ and Boyle CA (1998) Paternal dioxin and the sex of
children fathered by veterans of Operation Ranch Hand. Epidemiology 9,
Mocarelli P, Brambilla P, Gerthoux PM, Patterson DG, Jr and Needham LL
(1996) Change in sex ratio with exposure to dioxin. Lancet 348,409.
Mocarelli P, Gerthoux PM, Ferrari E et al. (2000) Paternal concentrations of
dioxin and sex ratio of offspring. Lancet 355,1858–1863.
Møller H (1998) Trends in sex-ratio, testicular cancer and male reproductive
hazards: are they connected? APMIS 106,232–238.
Parazzini F, La Vecchia C, Levi F and Franceschi S (1998) Trends in male:
female ratio among newborn infants in 29 countries from five continents.
Hum Reprod 13,1394–1396.
Richthoff J, Rylander L, Jo ¨nsson BA, A˚kesson H, Hagmar L, Nilsson-Ehle
P, Stridsberg M and Giwercman A (2003) Serum levels of 2,20,4,40,5,50-
hexachlorobiphenyl (CB-153) in relation to markers of reproductive func-
tion in young males from the general Swedish population. Environ Health
Rignell-Hydbom A, Rylander L, Giwercman A, Jo ¨nsson BA, Nilsson-Ehle P
and Hagmar L (2004) Exposure to CB-153 and p,p0-DDE and male repro-
ductive function. Hum Reprod 19,2066–2075.
Rosenfeld CS, Grimm KM, Livingston KA, Brokman AM, Lamberson WE
and Roberts RM (2003) Striking variation in the sex ratio of pups born to
mice according to whether maternal diet is high in fat or carbohydrate.
Proc Natl Acad Sci USA 100,4628–4632.
Rylander L and Hagmar L (1995) Mortality and cancer incidence among
women with a high consumption of fatty fish contaminated with persistent
organochlorine compounds. Scand J Work Environ Health 21,419–426.
Rylander L and Hagmar L (1999) No evidence for congenital malformations
or prenatal death in infants born to women with a high dietary intake of
fish contaminated with persistent organochlorines. Int Arch Occup Environ
Rylander L, Stro ¨mberg U and Hagmar L (1995) Decreased birth weight in
infants born to women with a high dietary intake of fish contaminated
with persistent organochlorine compounds. Scand J Work Environ Health
Schultz R, Suominen J, Varre T, Hakovirta H, Parvinen M, Toppari J and
Pelto-Huikko M (2003) Expression of aryl hydrocarbon receptor and aryl
hydrocarbon receptor nuclear translocator messenger ribonucleic acids and
proteins in rat and human testis. Endocrinology 144,767–776.
Sjo ¨din A, Hagmar L, Klasson-Wehler E, Bjo ¨rk J and Bergman A (2000)
Influence of the consumption of fatty Baltic Sea fish on plasma levels of
halogenated environmental contaminants in Latvian and Swedish men.
Environ Health Perspect 108,1035–1041.
Svensson BG, Nilsson A, Jo ¨nsson E, Schutz A, A˚kesson B and Hagmar L
(1995) Fish consumption and exposure to persistent organochlorine com-
pounds, mercury, selenium and methylamines among Swedish fishermen.
Scand J Work Environ Health 21,96–105.
Toppari J, Larsen JC, Christiansen P et al. (1996) Male reproductive health
and environmental xenoestrogens. Environ Health Perspect 104 (Suppl 4),
van der Pal-de Bruin KM, Verloove-Vanhorick SP and Roeleveld N (1997)
Change in male:female ratio among newborn babies in Netherlands.
WHO (1999) WHO laboratory Manual for Examination of Human Semen
and Sperm-Cervical Mucus Interaction. 4th edn. CUP, Cambridge.
Submitted on December 3, 2004; resubmitted on February 10, 2005;
accepted on February 22, 2005
POPs and sperm sex chromosome ratio