Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
RESEARCH
Open Access
Association of maternal serum concentrations of
2,2’, 4,4’5,5’-hexachlorobiphenyl (CB-153) and
1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene
(p,p’-DDE) levels with birth weight, gestational
age and preterm births in Inuit and European
populations
Bogdan J Wojtyniak1*, Daniel Rabczenko1, Bo AG Jönsson2, Valentyna Zvezday3, Henning S Pedersen4,
Lars Rylander2, Gunnar Toft5, Jan K Ludwicki6, Katarzyna Góralczyk6, Anna Lesovaya3, Lars Hagmar2,
Jens Peter Bonde5, the INUENDO research group7
Abstract
Background: Epidemiological studies on the association between maternal exposure to persistent organic
pollutants (POPs) and fetal growth alteration report inconsistent findings which weights in favor of additional
studies.
Methods: Blood samples were collected from interviewed pregnant women in Greenland (572), Kharkiv (611) and
Warsaw (258) and were analyzed for CB-153 and p,p’-DDE by gas chromatography-mass spectrometry. Data on
birth weight, gestational age and preterm birth were obtained for 1322 singleton live births. We examined the
association between natural log-transformed serum POPs concentration and birth weight and gestational age
using multiple linear regression and the association with prematurity using logistic regression controlling for
potential confounding factors.
Results: The median serum concentrations of CB-153 and p,p’-DDE were for Inuit mothers 105.6 and 298.9, for
Kharkiv mothers 27.0 and 645.4 and for Warsaw mothers 10.7 and 365.2 ng/g lipids, respectively. Increase in CB-153
concentration by one unit on the log scale in Inuit mothers serum was associated with significant decrease in
infant birth weight of -59 g and gestational age by -0.2 week. Decreases observed in the cohorts in Kharkiv (-10 g
and -0.1 week) and in Warsaw (-49 g and -0.2 week) were not statistically significant. Increase in p,p’-DDE
concentration by one unit on the log scale was associated with a statistically significant decrease in infant birth
weight of -39.4 g and -104.3 g and shortening of gestational age of -0.2 week and -0.6 week in the Inuit and
Warsaw cohorts, respectively. In the Kharkiv cohort decrease in birth weight (-30.5 g) was not significant, however
a shortening of gestational age of -0.2 week per increase in p,p’-DDE concentration by one unit on the log scale
was of the borderline significance. There was no significant association between CB-153 and p,p’-DDE
concentrations and risk of preterm birth however, in all cohorts the odds ratio was above 1.
Conclusions: In utero exposure to POPs may reduce birth weight and gestational age of newborns however, new
insights as to why results vary across studies were not apparent.
* Correspondence: bogdan@pzh.gov.pl
1
Department-Centre of Monitoring and Analyses of Population Health,
National Institute of Public Health - National Institute of Hygiene, Warsaw,
Poland
Full list of author information is available at the end of the article
© 2010 Wojtyniak et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
Background
Persistent Organic Pollutants (POPs) are chemicals that
persist in the environment, accumulate in high concentrations in fatty tissues and are bio-magnified through
the food-chain. They constitute a potential environmental hazard causing possible long-term risks to human
health. They are border-crossing substances transported
long distance, predominately by air and critical concentrations have been reached in some regions, even in
places where they have never been produced or used.
POPs have been detected in human blood, adipose tissue and breast milk all over the world. Chlorinated
hydrocarbon pesticides are not presently recommended
for the use in agriculture, and in public health programs
only a few persistent chlorinated hydrocarbon pesticides
are still being used in a few countries, especially DDT,
in vector disease control. Similarly, PCBs widely used by
the industry in the past and now ultimately banned for
at least two decades are still transported through different environmental media.
Although there has been a growing concern about the
effect of POPs on human fetal development and birth
outcomes, variation in results across studies remains
unexplained. Epidemiological findings regarding association of POPs such as polychlorinated biphenyls (PCBs)
and pesticides such as DDT with human fetal growth
alteration are inconsistent [1-4]. There are several studies demonstrating that exposure to PCBs at ordinary
environmental levels was related to a reduction in infant
birth weight [5-8] however, there are also studies showing little evidence to support such an association [9-12].
In all those studies PCB exposure was assessed from
maternal or cord blood. Similar inconsistency is present
in the studies with respect to gestational age or premature births [3,11-14]. Not less inconsistent are the
results concerning an association between exposure to
the major, persistent, and antiandrogenic DDT metabolite 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (p,p’DDE) and birth weight and length of gestation
[11,15,16].
The present study stems from a collaborative research
project funded by EU (Inuendo; http://www.inuendo.dk)
aiming to enlighten the impact of dietary POPs on
human reproductive function in an epidemiological setting of varying POPs exposure. The Inuendo project
study population has been established in three European
countries - Sweden, Poland (Warsaw) and Ukraine
(Kharkiv) together with a population of Inuits from
Greenland. Since we were able to follow-up most of the
participating pregnant women from Greenland, Warsaw
and Kharkiv till pregnancy termination it gave us an
opportunity to analyze some characteristics of the newborns as an additional measurements of POPs impact
Page 2 of 10
on human reproduction. We have chosen to use the
PCB congener, 2,2’, 4,4’5,5’-hexachlorobiphenyl (CB153) in serum as a biomarker for POPs exposure
because of its very high correlations with the total PCB
concentration [17-19]. Another relevant biomarker is
p,p’-DDE. Detailed analysis of the inter-population variation in CB-153 and p,p’-DDE serum concentrations and
of the determinants of this variation in the study populations has been given elsewhere [20]. The aim of this
study was to investigate the association between maternal serum level of CB-153 and p,p’-DDE and birth
weight, length of gestation and risk of preterm birth.
The number of low birth weight (below 2500 g) infants
was too small for a separate analysis of the risk of this
pregnancy outcome.
Methods
Between June 2002 and May 2004 we recruited pregnant
women and their male spouses in Greenland, Kharkiv
(Ukraine) and Warsaw (Poland) for interviews and
blood sampling. A general criterion for eligibility was
that the participants were born in the country of study
and were at least 18 years of age. A detailed description
of the recruitment process, population characteristics
and data collection has been given elsewhere [20,21]. In
brief, the target population consisted of pregnant
women who visited antenatal health care units. In
Greenland Inuit women lived in 19 municipalities and
settlements throughout the country, out of 665 eligible
approached 35 refused and 32 were inaccessible. Thus
598 (89.9%) were interviewed and 572 (86.0%) also
donated a blood sample. In Ukraine the women lived in
Kharkiv and surrounding villages and visited one of
three maternity hospitals or eight antenatal clinics in
the city. Altogether 2478 pregnant women were
informed about the project and asked to participate of
whom 632 (25.8%) were interviewed and 611 (24.7%)
donated a blood sample. One of the reasons that a large
number of pregnant women refused to participate was
concern that the collection of blood sample would
imply a risk for the pregnancy and the baby - in particular when anemia had been detected during pregnancy.
Demographic and reproductive information obtained
from 605 of those women who declined participation in
the study show that the average age in the group (22.8
years) was slightly lower than among those who participated (24.9 years), while the average number of children
in both the groups was similar (1.1 versus 1.2 among
those with at least one child). In Poland the women
lived mostly in Warsaw and its suburbs and visited the
obstetric out-patient clinic of a large Gynecological and
Obstetric Hospital or physicians at a collaborating hospital also in Warsaw. Altogether 690 women were
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
Page 3 of 10
informed about the project and invited to participate of
whom 472 (68.4%) were interviewed and 258 (37.4%)
donated a blood sample. In total 1441 women were
interviewed and donated blood and 1322 of them delivered a singleton infant for whom all required information (gestational age, sex and birth weight) were
available. Analyzed sample size, pregnancy outcomes
and levels of exposure chemicals are presented in Table
1 and Figure 1.
The Inuit and Ukrainian women were earlier in their
pregnancies at the time of interview and blood sample
collection (on average 24 weeks pregnant) than the Polish women (33 weeks pregnant). Through the interview
we obtained information on the maternal demographic
and social factors and time varying characteristics such
as weight, smoking, alcohol drinking and occupational
exposures were obtained for the period of attempt to
conceive (Table 2).
The study was approved by local ethical committees
and all subjects signed an informed consent.
Determination of CB-153 and p,p’-DDE in serum
All analysis of CB-153 and p,p’-DDE in serum were performed in the Department of Occupational and Environmental Medicine in Lund, Sweden and the procedures
were described in detail elsewhere [20,22,23]. Briefly,
serum concentrations of CB-153 and p,p’-DDE were
analyzed by gas chromatography mass spectrometry following solid phase extraction. The relative standard
deviations, calculated from samples analyzed in duplicate at different days, was 18% at 0.1 ng/mL (n = 990),
10% at 0.5 ng/mL (n = 990) and 10% at 2 ng/mL (n =
990) for CB-153 and 11% at 1 ng/mL (n = 1058), 8% at
3 ng/mL (n = 1058) and 7% at 8 ng/mL (n = 1058) for
p,p’-DDE.
The determination limits were 0.05 ng/mL for CB-153
and 0.1 ng/mL for p,p’-DDE. In the case of non-detectable concentrations, the results were set at half of the
Table 1 Analyzed sample size, levels of pregnancy
outcomes and of exposure chemicals
Greenland
Kharkiv
Warsaw
Mothers approached
665 (100.0%)
2478 (100.0%)
690 (100.0%)
Mothers interviewed
Blood samples
598 (89.9%)
572 (95.7%)
640 (25.8%)
611 (95.5%)
472 (68.4%)
258 (54.7%)
CB-153 [ng/g lipid]
1
p,p’-DDE [ng/g lipid]
1
Live births analyzed
Preterm births
Gestational age [weeks]
Birth weight [grams]
1
geometric mean
full term births
2
2
105.4 (± 2.8)
25.7 (± 1.9)
9.0 (± 2.1)
273.8 (± 2.9)
653.3 (± 1.8)
356.8 (± 1.9)
547 (82.3%)
577 (23.3%)
198 (28.7%)
28 (5.1%)
12 (2.1%)
12 (6.1%)
39.6 (± 1.9)
39.1 (± 1.2)
39.2 (± 1.8)
3582 (± 611)
3273 (± 438)
3453 (± 506)
determination limit. There were 69 such values for CB153 and 10 for p,p’-DDE. The results were expressed on
a lipid weight basis determined by enzymatic methods
[20].
Outcome measures
At the end of the field study the pregnancy outcome
short questionnaire was filled in by medical personnel at
the health centers where delivery took place using data
from personal forms of pregnant women and the history
of their delivery. Three outcomes were assessed in the
analysis: 1) birth weight (in grams) analyzed only for
full-term birth - 37 and more weeks of completed gestation; 2) gestational age (in weeks) defined as the duration between a self-reported date of the last menstrual
period and the date of birth; and 3) preterm birth,
before 37 week of gestation. The information from singleton live births were taken into account only. Of 1322
births 52 (3.9%) were born preterm while only 19 full
term newborns had low birth weight (less than 2500 g)
therefore we did not analyze this outcome variable. The
mothers who were interviewed but were not included in
the birth outcome analysis (26 in Greenland, 29 in
Kharkiv and 214 in Warsaw) did not differ significantly
from the analyzed group taking into account the sociodemographic variables that were considered as potential
confounding factors.
Statistical analysis
We examined the association between CB-153 and p,p’DDE serum concentrations and birth weight and gestational age, respectively, using linear regression models
and relation to prematurity using logistic regression
models built for each country independently.
We assessed the shape of the relationship between
POPs concentrations and outcome variables using generalized additive models with integrated smoothness
estimation and derived number of the effective degrees
of freedom (EDF) [24]. If the number of EDF was above
one (thus relationship could be viewed as non-linear)
we fitted a parametric model with natural log-transformed POP variable. Nonlinear association was found
in 10 of 18 analyzed relationships. Next, using analysis
of deviance [25], we compared fit of non-parametric and
logarithmic models. In all the cases there were no significant differences between the fit. If dependence could be
regarded as linear (EDF = 1) we checked if a linear
model could be replaced by a logarithmic one. The
comparison was based on R 2 in the case of ordinary
regression and in terms of a generalized R2 (computed
as a ratio of deviance of the model to null deviance) in
the case of logistic regression. Since values of the compared statistics were similar, we report estimates from
the logarithmic model for easier comparability of the
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
Page 4 of 10
Figure 1 Distribution of maternal serum concentration of CB-153 and p,p’-DDE in the study populations (left-hand side axis has a
logarithmic scale for untransformed concentration values, right-hand side scale is linear for log-transformed concentration values).
results. The effect point estimates are presented together
with a 95% confidence interval (95% C.I.). Because analyzed concentrations of CB-153 and p,p’-DDE were logtransformed linear regression coefficients presented in
the tables represent changes in the outcome variables
(birth weight, gestational age) associated with increase
in chemical concentration by one unit on the log scale.
In the logistic regression analysis that we applied to premature birth, the effects of concentrations are presented
as odds ratios associated with their twofold increase. In
the case of natural-log transformed independent variable
“x” odds ratio associated with change of this dependent
variable from its value x 1 to x 2 is given by formula
were included in the models in a stepwise forward fashion if the addition of each variable to the model changed the regression coefficient of the chemical serum
concentration by at least 10% [26]. In the final step of
the birth weight model we added gestational age, known
to influence birth weight, to check to what extent birth
weight reduction due to the POP exposure may be
related to shortened gestational age.
The analyses were done using the S-PLUS 2000 [27]
and R 2.9.2 [28] statistical programs. The term statistically significant in all performed analyses indicates a pvalue of less than 0.05.
x 2 where b1 is a regression coefficient. To show
e
results in a generalized way, we applied common
approach in such a situation and calculated odds ratio
for the case when x1/x2 = 2 which is denoted as odds
ratio associated with twofold increase of untransformed
concentration.
To control for the confounding of the exposure-outcome association by other factors we built multivariate
models taking into account such covariates as maternal
age, body mass index (BMI) before pregnancy, education, marital status, smoking status, alcohol drinking,
occupational exposure (to paints, solvents, fumes, engine
exhaust), parity and also the newborn’s sex. Covariates
Results
There were differences in the mean birth weight of singleton, live, full term newborns among the study cohorts
(Table 1). The heaviest were newborns in Greenland and
the lightest in Kharkiv. On the other hand preterm births
were less frequent in Kharkiv than in the other two study
groups. The Inuit mothers had a distinctly higher serum
concentration of CB-153 as compared with mothers from
Warsaw and Kharkiv (age-adjusted geometric mean ratio
12.7 and 3.8, respectively). The highest serum concentrations of p’p-DDE was observed in Kharkiv mothers (ageadjusted geometric mean ratio 2.52 and 2.02 when compared to Inuit and Warsaw mothers, respectively).
x
b1 ln( 1 ) ,
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
Page 5 of 10
Table 2 Characteristics of the study cohorts - potential confounding factors
Greenland
Kharkiv
Warsaw
Age [years]
26.8 (± 6.1)
6.1 (± 25)
25 (± 4.8)
Pre-pregnancy BMI [kg/m2]
24.5 (± 4.3)
4.3 (± 21.6)
21.6 (± 3.2)
227 (41.5%)
156 (27%)
-
46 (8.4%)
144 (25%)
44 (22.2%)
continued above age 18
211 (38.6%)
207 (35.9%)
150 (75.8%)
missing
63 (11.5%)
70 (12.1%)
4 (2.0%)
married
114 (20.8%)
322 (55.8%)
160 (80.8%)
living as married
living alone
323 (59.0%)
55 (10.1%)
195 (33.8%)
51 (8.8%)
32 (16.2%)
6 (3.0%)
Education
left before 15 or 15-17
left at age 18
Marital status
Smoking1
Exposure to other people’s cigarette smoke1
Alcohol
drinks1
missing
55 (10.1%)
9 (1.6%)
0 (0.0%)
never smoker
64 (11.7%)
368 (63.8%)
129 (65.2%)
smoker
65 (11.9%)
78 (13.5%)
32 (16.2%)
ex-smoker
418 (76.4%)
131 (22.7%)
37 (18.7%)
no
146 (26.7%)
110 (19.1%)
71 (35.9%)
yes
missing
390 (71.3%)
11 (2.0%)
455 (78.9%)
12 (2.1%)
126 (63.6%)
1 (0.5%)
less than 14/week
519 (94.9%)
576 (99.8%)
197 (99.5%)
14 and more/week
Occupational
exposure1
no
yes
Previous pregnancies
0
1 (0.2%)
1 (0.5%)
497 (86.1%)
172 (86.9%)
8 (1.5%)
80 (13.9%)
26 (13.1%)
63 (11.5%)
320 (55.5%)
148 (74.7%)
1
83 (15.2%)
127 (22.0%)
34 (17.2%)
2+
395 (72.2%)
107 (18.5%)
9 (4.5%)
6 (1.1%)
23 (4.0%)
7 (3.5%)
295 (53.9%)
292 (53.4%)
300 (52.0%)
277 (48.0%)
96 (48.5%)
102 (51.5%)
missing
Newborn sex
28 (5.1%)
539 (98.5%)
Boy
Girl
1
during the period of attempt to conceive
Overall, we found statistically significant relations
between maternal serum concentration of CB-153 and
birth weight and length of gestation only in the Inuit
population (Table 3).
Increase in log CB-153 concentration was associated
with statistically significant decrease in infants birth
weight, however adjustment for confounding variables in
a multivariate models reduced the effect preserving its
significance.
It indicates that the possible influence of CB-153
exposure on birth weight of Inuit infants is in part
mediated through a change in gestational age. This
sequence is further corroborated by a significant albeit
small reduction in gestational age in Inuit infants associated with increased CB-153 concentration in maternal
blood. However, even in the Inuit population the risk of
premature birth did not grow significantly with the
increase in the serum concentration of CB-153. We did
not observe a statistically significant association between
maternal serum CB-153 concentration and infant birth
weight, gestational age or risk of prematurity in Kharkiv
and Warsaw. It may be stressed, however, that in all
cohorts the association was inverse and the adjusted
odds ratios of prematurity associated with the increase
in log CB-153 concentration were greater than 1.
The results of analysis for p,p’-DDE maternal serum
concentration and birth weight and gestational age were
more conspicuous than those for CB-153 and we
observed statistically significant associations in Inuit and
in the Warsaw cohorts. In an unadjusted model increase
in log p,p’-DDE concentration was associated with a statistically significant decrease in infant birth weight
(Table 4). Control for confounders reduced the effect,
but it retained its significance. Adding gestational age
reduced the effect by one fourth, nevertheless the effect
in Warsaw was significant, and in Greenland was of borderline significance.
The gestational age was significantly reduced when
maternal serum p,p’-DDE concentration was increasing
in unadjusted and multivariate models in Greenland and
Warsaw. In the Kharkiv cohort the gestational age
reduction associated with an increase in log p,p’-DDE
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
Page 6 of 10
Table 3 Regression coefficients1 and odds ratios2 (OR) from unadjusted and multivariate linear and logistic regression
models of maternal serum log CB-153 and birth weight, gestational age and premature birth
Birth weight (grams)
Regression
95% C.I.
Gestational age (weeks)
p
coefficient
Regression
Prematurity
95% C.I.
p
OR
95% C.I.
p
0.16
coefficient
Unadjusted3 model
Greenland
-87.6
(-132.9; -42.3)
< 0.01
-0.2
(-0.4; 0.0)
0.02
1.21
(0.93; 1.58)
Kharkiv
-24.6
(-75.2; 26.0)
0.34
-0.2
(-0.4; 0.0)
0.05
1.43
(0.78; 2.63)
0.26
Warsaw
-45.6
(-129.1; 37.9)
0.29
-0.1
(-0.5; 0.3)
0.62
0.93
(0.55; 1.58)
0.79
Greenland
-72.7
(-118.0; -27.4)
< 0.01
-0.2
(-0.4; 0.0)
0.02
1.14
(0.86; 1.52)
0.36
Kharkiv
-17.9
(-71.2; 35.4)
0.51
-0.1
(-0.3; 0.1)
0.12
1.28
(0.62; 2.65)
0.51
Warsaw
-61.6
(-152.3; 29.1)
0.19
-0.2
(-0.6; 0.2)
0.40
1.29
(0.65; 2.55)
0.47
Multivariate model3,4
Multivariate model, with gestational age
Greenland
-59.2
(-100.6; -17.8)
Kharkiv
-10.2
(-61.4; 41.0)
0.01
0.7
Warsaw
-49.4
(-134.5; 35.7)
0.26
1
Per one unit increase of natural log CB-153
Per twofold increase of untransformed CB-153
3
Number of subjects in unadjusted/multivariate models: Greenland birth weight 519/519, gestational age 547/547, prematurity 547/547; Kharkiv birth weight 565/
554, gestational age 577/577, prematurity 577/577; Warsaw birth weight 186/186, gestational age 198/198, prematurity 198/198
4
Models for birth weight controlled for: smoking status (Greenland); age of mother, mother’s BMI, gender of child, number previous pregnancies, mother’s martial
status, passive smoking (Kharkiv); age of mother, gender of child, mother’s education status, mother’s BMI (Warsaw);
Models for gestational age controlled for: none (Greenland); mother’s age (Kharkiv); mother’s education level, mother’s age, previous pregnancies, exposition,
alcohol drinking (Warsaw);
Models for prematurity controlled for: mother’s education level, mother’s age, martial status (Greenland); mother’s education level, previous pregnancies, smoking
status, marital status, mother’s age (Kharkiv); mother’s BMI, mother’s age (Warsaw);
2
Table 4 Regression coefficients1 and odds ratios2 (OR) from unadjusted and multivariate linear and logistic regression
models of maternal serum log p,p’-DDE and birth weight, gestational age and premature birth
Birth weight (grams)
Regression
95% C.I.
Gestational age (weeks)
p
coefficient
Regression
Prematurity
95% C.I.
p
OR
95% C.I.
p
coefficient
Unadjusted3model
Greenland
-71.4
(-114.7; -28.1)
< 0.01
-0.2
(-0.4; 0.0)
0.02
1.13
(0.87; 1.47)
0.37
Kharkiv
-32.2
(-90.8; 26.4)
0.28
-0.2
(-0.4; 0.0)
0.09
1.43
(0.72; 2.82)
0.31
Warsaw
-131.3
(-231.5; -31.1)
0.01
-0.5
(-0.9; -0.1)
0.02
1.45
(0.73; 2.88)
0.30
Multivariate model3,4
Greenland
-56
(-99.5; -12.5)
0.01
-0.2
(-0.4; 0.0)
0.04
1.07
(0.81; 1.41)
0.64
Kharkiv
-48.9
(-109.7; 11.9)
0.12
-0.2
(-0.4; 0.0)
0.06
1.60
(0.75; 3.44)
0.23
Warsaw
-146.9
(-248.8; -45.0)
0.01
-0.6
(-1; -0.2)
< 0.01
2.44
(0.99; 6.06)
0.05
Multivariate model, with gestational age
Greenland
-39.4
(-79.0; 0.2)
Kharkiv
-30.5
(-88.7; 27.7)
0.3
Warsaw
-104.3
(-202.5; -6.1)
0.04
1
0.05
Per one unit increase of natural log p,p’-DDE
Per twofold increase of untransformed p,p’-DDE
3
Number of subjects in unadjusted/multivariate models: Greenland birth weight 519/519, gestational age 547/547, prematurity 547/547; Kharkiv birth weight 565/
565, gestational age 577/577, prematurity 577/577; Warsaw birth weight 186/186, gestational age 198/198, prematurity 198/198
4
Models for birth weight controlled for: smoking status (Greenland); age of mother (Kharkiv); age of mother, mother’s BMI (Warsaw);
Models for gestational age controlled for: passive smoking (Greenland); mother’s age (Kharkiv); mother’s education (Warsaw);
Models for prematurity controlled for: mother’s education level, mother’s age, mother’s BMI, marital status, passive smoking (Greenland); mother’s age (Kharkiv);
education level, smoking status, marital status, mother’s age (Warsaw);
2
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
exposure was of similar magnitude as in the Greenland
cohort, however at the borderline of significance.
We observed no significant association between
maternal serum concentrations of p,p’-DDE and the risk
of premature birth in any study cohort however, all the
odds ratios were above 1 and in Warsaw at the borderline of significance.
A very high correlation between the two POP biomarkers in the Inuit study group (Spearman’s correlation
coefficient rS = 0.92) made impossible to disentangle the
independent effect of each of them. In Warsaw and
Kharkiv the weaker correlation (r S = 0.55 and 0.49
respectively) allowed us to run models for birth weight
and gestational age with both compounds. Only the
association between p,p’-DDE and gestational age in the
Warsaw cohort remained statistically significant (-0.7
week decrease, 95% C.I. -1.2 - -0.2, p < 0.01).
Discussion
Epidemiological evidence regarding adverse effects of
POPs on human pregnancy outcome is still limited and
results supporting a detrimental association go together
with observations that fail to corroborate these results.
Our study found a significant reduction in birth weight
and length of gestation associated with maternal CB-153
exposure among Inuits but not in the Kharkiv or Warsaw cohorts. However, maternal p,p’-DDE exposure was
significantly associated with a reduction in birth weight
in the Inuit as well as the Warsaw cohorts, and in the
case of gestational age we found the association in all
three cohorts (in Kharkiv at the borderline of
significance).
Our study found no significant effect of CB-153 and
p,p’-DDE on the risk of prematurity however, in all
cohorts the OR was above one (in Poland at the borderline of significance). It should be stressed, that small
numbers of preterm births, 12 in each of the Warsaw
and Kharkiv cohorts and 28 in the Inuit cohort, definitely limited statistical power of our analysis.
As could be expected serum concentrations of CB-153
were much higher in the Inuit mothers as compared
with Warsaw and Kharkiv mothers, whereas p,p’-DDE
concentrations were highest in the Kharkiv mothers and
were rather similar in the Inuit and Warsaw mothers.
The exposure pattern in the Kharkiv mothers is in
accordance with a previous analysis of maternal milk
from Ukraine in which the median p,p’-DDE concentration (2457 ng/g lipid) was one order of magnitude
higher than the median CB-153 concentration (149 ng/g
lipid) [29]. Concentration levels of CB-153 in the Kharkiv and Warsaw cohorts and of p,p’-DDE in the Inuit
and Warsaw cohorts were similar to the concentrations
observed in US female population from the National
Health and Nutrition Examination Survey in 2003-2004
Page 7 of 10
(median 21.9 ng/g lipid and 207 ng/g lipid for CB-153
and p,p’-DDE, respectively) [30]. It may be of interest to
mention that the level of CB-153 concentration in Warsaw and Kharkiv mothers was much lower than
observed in the mothers of two districts in neighboring
Slovakia (median 140 ng/g lipid) [31].
The possible explanation why the association of CB153 with birth weight and gestational age was found
only among the Inuits may be a much higher exposure
level in the Inuit cohort than in the other two. In previous studies, where significant association was
observed, the exposure levels due to dietary intakes had
been similar or higher than in our Inuit cohort [5-8,13]
or were high due to occupational exposure [32], or accidental poisoning [33,34]. On the other hand, however,
in some populations, where the exposure levels were
similar or higher than among the Inuits in the present
study, no significant associations were found
[3,9,11,31,35,36]. In a rural Spanish population [10]
where the PCB exposure level was similar to that
observed in Warsaw and in the population of two Ukrainian cities other than Kharkiv [29] maternal PCB was
also not associated with birth weight. Thus, it seems
reasonable to conclude that maternal exposure to CB153 at the levels observed in Warsaw and Kharkiv is not
associated with detrimental effects on newborns birth
weight, gestational age or risk of prematurity.
Association of maternal DDE exposure with the
decrease of birth weight that we observed in the Inuit
and Warsaw cohorts corroborates earlier findings in a
Great Lakes population study [11], in the Collaborative
Perinatal Project (CPP) [15] and in a study in India [37].
On the other hand the lack of association observed in
the Kharkiv cohort is in agreement with the results of
several other studies that found no association of DDE
with birth weight [8-10,16,35,38]. The consistent reverse
association of the maternal DDE exposure and the
length of gestation that we observed in all three study
cohorts (in Kharkiv at the borderline of significance)
substantiates the earlier results of the CPP [15].
To take into account the differences in the level of
CB-153 as well as p,p’-DDE in the study cohorts we calculated changes in the outcome variables associated
with the interquartile change of both POP biomarkers
in each study cohort. Such a change is a good measure
of the public health importance of the variation in
population exposure. An interquartile increase in concentration of CB-153 was associated with the reduction
of birth weight by 74.5 g, 8.1 g and 49.4 g, the reduction
of gestational age by 0.3, 0.1 and 0.2 weeks and the
increase of odds of prematurity by 2.01, 2.26 and 2.85
times in the Inuit, Kharkiv and Warsaw cohorts, respectively. An interquartile increase in concentration of p,p’DDE was associated with the reduction of birth weight
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
by 47.6 g, 22.9 g and 80.0 g, the reduction of gestational
age by 0.3, 0.2 and 0.5 weeks and the increase of odds
of prematurity by 1.39, 4.35 and 17.39 times in the
Inuit, Kharkiv and Warsaw cohorts, respectively.
The results from the Inuit cohort in Greenland speak
in favor of an inverse association between POPs exposure and birth weight and gestational age. However, the
high correlation between CB-153 and p,p’-DDE concentrations in serum made it impossible to disentangle
independent effects from these compounds. It should
also be noted that both CB-153 and p,p’-DDE act as
index biomarkers, and other POP compounds could be
the ones which affect fetal growth. However, it is of
interest to note, that a significant association between p,
p’-DDE and a reduction in birth weight and gestational
age, was found also in the Warsaw cohort, in which the
correlation between CB-153 and p,p’-DDE was only
moderate. Moreover, the association in Warsaw was
stronger than in Greenland. In the case of gestational
age this association retained its significance when we
also controlled for CB-153. In Kharkiv both CB-153 and
p,p’-DDE were non-significant when simultaneously
included in the models. We have not identified a study
that would demonstrate an inverse independent association of PCBs and DDE with birth weight or gestational
age.
Our results provide only limited evidence that
increased CB-153 or p,p’-DDE levels are associated with
an increased risk of preterm birth. In all three cohorts
and for both chemicals the risks of prematurity were
elevated but only for CB-153 concentration in the Warsaw cohort the association reached statistical significance at borderline level. As mentioned above small
numbers of preterm births limited statistical power of
our analysis. Results of other studies are inconsistent
demonstrating presence of significant association
[10,15,39] as well as the lack of it [3,16,40,41].
Our study has several strengths. In all participating
countries the study was set-up and carried out according to an agreed uniform research protocol regarding
measurement of exposure and pregnancy outcome. We
used maternal serum CB-153 and p,p’-DDE concentrations as index biomarkers representing direct measurement of POPs as the indicator of intrauterine exposure.
Determination of p,p’-DDE and CB-153 in maternal
serum was performed for all samples in one specialized
center.
On the other hand our study has some limitations as
well. The participation rate in the study varied considerably between the populations and while in Greenland
we obtained full information from more than 80% of
those invited, in Kharkiv the participation rate was only
25%. We have no reason to suspect that the mothers’
decision to participate was associated with knowledge of
Page 8 of 10
POPs exposure or expected pregnancy outcome in
Greenland or in Warsaw. In Kharkiv however, low risk
of premature birth (2.1%) could be a result of collecting
part of the data (about 12%) at maternity hospitals
which were not taking care of mothers with previous
preterm birth since such mothers received care at a specialized neonatal hospital. In a study of pregnancy outcome in two urban areas of Ukraine the rate of preterm
delivery was 6.6% of singleton live births [42]. This
selection bias may be responsible for differing results in
the Kharkiv cohort. Also the Warsaw cohort was
selected to some extent because it mostly comprised
mothers who attended maternity training organized by
the hospital. We compared some key variables (birth
weight, mothers education, length of gestation, prematurity) in the study cohort and in all births in Warsaw
in 2003 that we obtained from the city register. This
comparison shows that although our study mothers had
on average better education the other variables were not
different. We were not able to obtain such data for
Kharkiv or Greenland.
Conclusions
Our results provide some epidemiologic support for an
association between in utero exposure to the POP biomarkers CB-153 and p,p’-DDE and reduction in birth
weight and gestational age however, new insights as to
why results vary across studies are not apparent.
Because observed birth weight reduction was independent of gestational age our findings suggest detrimental
effect of the exposure on intrauterine growth as well as
on length of gestation.
Abbreviations
BMI: body mass index; LOG: natural logarithm; PCB: polychlorinated
biphenyls; CB-153: 2,2’, 4,4’5,5’-hexachlorobiphenyl; p,p’-DDE: 1,1-dichloro-2,2bis (p-chlorophenyl)-ethylene; DDT: dichlorodiphenyl trichloroetane; POP:
persistent organic pollutant; EDF: effective degrees of freedom.
Acknowledgements
This study was partly supported by grants from the European Commission
(QLK4-CT-2001-00202), Swedish Research Council (Grant No: 521-2004-6072),
the Swedish Research Council for Environment, Agricultural Sciences and
Spatial Planning.
The authors thank Ms Hélène Åkesson, Ms Berit Holmskov, and Ms Christina
Held for performing chemical analyses in a skilful way.
The authors would like to thank the reviewers: Dr. Matthew Longnecker and
Dr. Michelle Mendez for their valuable comments that helped to improve
the manuscript.
Author details
1
Department-Centre of Monitoring and Analyses of Population Health,
National Institute of Public Health - National Institute of Hygiene, Warsaw,
Poland. 2Division of Occupational and Environmental Medicine, Lund
University, Lund, Sweden. 3Laboratory of Human Reproduction, Kharkiv State
Medical University, Kharkiv, Ukraine. 4Centre for Arctic Environmental
Medicine, Nuuk, Greenland. 5Department of Occupational Medicine, Aarhus
University Hospital, Aarhus, Denmark. 6Department of Environmental
Toxicology, National Institute of Public Health - National Institute of Hygiene,
Warsaw, Poland. 7www.inuendo.dk.
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
Authors’ contributions
JPB, LH designed and initiated Inuendo project. GT, HSP, JKL, KG, VZ and AL
were responsible for collecting the blood samples and the interview data.
BAGJ was responsible for the chemical analysis of the POP biomarkers. JPB
and GT coordinated the execution of the Inuendo project. GT has main
responsibility for creating Inuendo database. BJW initiated the newborn
study. DR and BJW had main responsibility for creating newborn database.
BJW and DR were responsible for statistical analysis and writing the draft
version of manuscript. All authors commented on and approved the final
manuscript.
Competing interests
The authors declare that they have no competing interests.
Received: 25 November 2009 Accepted: 6 September 2010
Published: 6 September 2010
References
1. 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. Annu Rev Public
Health 1997, 18:211-244.
2. Toft G, Hagmar L, Giwercman A, Bonde JP: Epidemiological evidence on
reproductive effects of persistent organochlorines in humans. Reprod
Toxicol 2004, 19:5-26.
3. Longnecker MP, Klebanoff MA, Brock JW, Guo X: Maternal Levels of
Polychlorinated Biphenyls in Relation to Preterm and Small-forGestational-Age Birth. Epidemiology 2005, 16:641-647.
4. Rogan WJ, Chen A: Health risks and benefits of bis(4-chlorophenyl)-1,1,1trichloroethane (DDT). Lancet 2005, 366:763-773.
5. Fein GG, Jacobson JL, Jacobson SW, Schwartz PM, Dowler JK: Prenatal
exposure to polychlorinated biphenyls: effects on birth size and
gestational age. J Pediatr 1984, 105:315-320.
6. Patandin S, Koopman-Esseboom C, de Ridder MA, Weisglas-Kuperus N,
Sauer PJ: Effects of environmental exposure to polychlorinated biphenyls
and dioxins on birth size and growth in Dutch children. Pediatr Res 1998,
44:538-545.
7. Rylander L, Stromberg U, Dyremark E, Ostman C, Nilsson-Ehle P, Hagmar L:
Polychlorinated biphenyls in blood plasma among Swedish female fish
consumers in relation to low birth weight. Am J Epidemiol 1998,
147:493-502.
8. Karmaus W, Zhu X: Maternal concentration of polychlorinated biphenyls
and dichlorodiphenyl dichlorethylene and birth weight in Michigan fish
eaters: a cohort study. Environmental Health, A Global Access Science Source
2004, 3:1.
9. Grandjean P, Bjerve KS, Weihe P, Steuerwald U: Birthweight in a fishing
community: significance of essential fatty acids and marine food
contaminants. Int J Epidemiol 2001, 30:1272-1278.
10. Ribas-Fitó N, Sala M, Cardo E, Mazón C, De Muga ME, Verdú A, Marco E,
Grimalt JO, Sunyer J: Association of Hexachlorobenzene and Other
Organochlorine Compounds with Anthropometric Measures at Birth.
Pediatric Research 2002, 52:163-167.
11. Weisskopf MG, Anderson HA, Hanrahan LP, Kanarek MS, Falk CM,
Steenport DM, Draheim LA, the Great Lakes Consortium: Maternal
exposure to Great Lakes sport-caught fish and dichlorodiphenyl
dichloroethylene, but not polychlorinated biphenyls, is associated with
reduced birth weight. Environ Res 2005, 97:149-62.
12. Longnecker MP: Invited commentary: Why DDT matters now. Am J
Epidemiol 2005, 162:726-728.
13. Hertz-Picciotto I, Charles MJ, James RA, Keller JA, Willman E, Teplin S: In
utero polychlorinated biphenyl exposures in relation to fetal and early
childhood growth. Epidemiology 2005, 16:648-56.
14. Fein GG, Jacobson JL, Jacobson SW, Schwartz PM, Dowler JK: Prenatal
exposure to polychlorinated biphenyls: effects on birth size and
gestational age. J Pediatr 1984, 105:315-320.
15. Longnecker MP, Klebanoff MA, Zhou HB, Brock JW: Association between
maternal serum concentration of the DDT metabolite DDE and preterm
and small-for-gestational-age babies at birth. Lancet 2001, 358:110-114.
16. Fahrang L, Weintraub JM, Petreas M, Eskenazi B, Bhatia R: Association of
DDT and DDE with Birth Weight and Length of Gestation in the Child
Page 9 of 10
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
Health and Development Studies, 1959-1967. Am J Epidemiol 2005,
162:717-725.
Glynn AW, Wolk A, Aune M, Atuma S, Zettermark S, Maehle-Schmid M,
Darnerud PO, Becker W, Vessby B, Adami HO: Serum concentrations of
organochlorines in men: a search for markers of exposure. Sci Total
Environ 2000, 263:197-208.
Grimvall W, Rylander L, Nilsson-Ehle P, Nilsson U, Strömberg U, Hagmar L,
Östman C: Monitoring of polychlorinated biphenyls in human blood
plasma with respect to age, lactation and fish consumption;
methodology developments. Arch Environ Contam Toxicol 1997,
32:329-336.
Longnecker MP, Ryan JJ, Gladen BC, Schecter AJ: Correlations among
human plasma levels of dioxin-like compounds and polychlorinated
biphenyls (PCBs) and implications for epidemiologic studies. Arch Environ
Health 2000, 55:195-200.
Jönsson BAG, Rylander L, Lindh C, Rignell-Hydbom A, Giwercman A, Toft G,
Pedersen HS, Ludwicki JK, Góralczyk K, Zvyezday V, Spanò M, Bizzaro D,
Bonefeld-Jörgensen EC, Manicardi GC, Bonde JP, Hagmar L, Inuendo: Interpopulation variations in concentrations, determinants of and
correlations between 2,2’,4,4’,5,5’-hexachlorobiphenyl (CB-153) and 1,1dichloro-2,2-bis (p-chlorophenyl)-ethylene (p,p’-DDE): a cross-sectional
study of 3161 men and women from Inuit and European populations.
Environmental Health: A Global Access Science Source 2005, 4:27.
Toft G, Hagmar L, Giwercman A, Bonde JP: Epidemiological evidence on
reproductive effects of persistent organochlorines in humans. Reprod
Toxicol 2004, 19:5-26.
Richthoff J, Rylander L, Jonsson BA, Akesson H, Hagmar L, Nilsson-Ehle P,
Stridsberg M, Giwercman A: Serum levels of 2,2’,4,4’,5,5’hexachlorobiphenyl (CB-153) in relation to markers of reproductive
function in young males from the general Swedish population. Environ
Health Perspect 2003, 111:409-413.
Rignell-Hydbom A, Rylander L, Giwercman A, Jönsson BA, Nilsson-Ehle P,
Hagmar L: Exposure to CB-153 and p,p’-DDE and male reproductive
function. Hum Reprod 2004, 19:2066-2075.
Wood S: Generalized Additive Models An Introduction with R, Chapman
and Hall/CRC. 2006.
Chambers JM, Hastie TJ: Statistical Models in S. Chapman and Hall
Computer Science Series 1993.
Moldanado G, Greenland S: Simulation study of confounder-selection
strategies. Am J Epidemiol 1993, 138:923-936.
MathSoft, Inc: S-PLUS 2000 User’s Guide, Seattle, Washington 1999.
R Development Core Team: R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria 2009
[http://www.R-project.org], ISBN 3-900051-07-0.
Gladen BC, Shkiryak-Nyzhnyk ZA, Chyslovska N, Zadorozhnaja TD, Little RE:
Persistent organochlorine compounds and birth weight. Ann Epidemiol
2003, 13:151-157.
Centers for Disease Control and Prevention (CDC): Fourth National Report
on Human Exposure to Environmental Chemicals. Atlanta (GA) 2009.
Sonneborn D, Park H-Y, Petrik J, Kacan A, Palkovicova L, Trnovec T,
Nguyen D, Hertz-Picciotto I: Prenatal polychlorinated biphenyl exposures
in eastern Slovakia modify effects of social factors on birthweight.
Paediatr Perinal Epidemiol 2008, 22:202-213.
Taylor PR, Stelma JM, Lawrence CE: The relation of polychlorinated
biphenyls to birth weight and gestational age in the offspring of
occupationally exposed mothers. Am J Epidemiol 1989, 129:395-406.
Yamashita F, Hayashi M: Fetal PCB syndrome: clinical features,
intrauterine growth retardation and possible alteration in calcium
metabolism. Environ Health Perspect 1985, 59:41-45.
Rogan WJ, Gladen BC, Hung KL, Koong SL, Shih LY, Taylor JS: Congenital
poisoning by polychlorinated biphenyls and their contaminants in
Taiwan. Science 1988, 241:334-336.
Rogan WJ, Gladen BC, McKinney JD, Carreras N, Hardy P, Thullen J,
Tinglestad J, Tully M: Neonatal effects of transplancental exposure to
PCBs and DDE. J Pediatr 1986, 109:335-341.
Vartiainen T, Jaakkola JJ, Saarikoski S, Tuomisto J: Birth weight and sex of
children and the correlation to the body burden of PCDDs/PCDFs and
PCBs of the mother. Environ Health Perspect 1998, 106:61-66.
Siddiqui MKJ, Nigam U, Srivastava S, Tejeshwar DS, Chandrawati :
Association of maternal blood pressure and hemoglobin level with
organochlorines in human milk. Hum Exp Toxicol 2002, 21:1-6.
Wojtyniak et al. Environmental Health 2010, 9:56
http://www.ehjournal.net/content/9/1/56
Page 10 of 10
38. Sagiv SK, Tolbert PE, Altshul LA, Korrick SA: Organochlorine Exposures
During Pregnancy and Infant Size at Birth. Epidemiology 2007, 18:120-128.
39. Wassermann M, Ron M, Bercovici B, Wassermann D, Cucos S, Pines A:
Premature delivery and organochlorine compounds: polychlorinated
biphenyls and some organochlorine insecticides. Environ Res 1982,
28:106-112.
40. Berkowitz GS, Lapinski RH, Wolff MS: The role of DDE and polychlorinated
biphenyl levels in preterm birth. Arch Environ Contam Toxicol 1996,
30:139-141.
41. Torres-Arreola L, Berkowitz G, Torres-Sánchez L, López-Cervantes M,
Cebrián ME, Uribe M, López-Carrillo L: Preterm birth in relation to
maternal organochlorine serum levels. Ann Epidemiol 2003, 13:158-162.
42. Monaghan SC, Little RE, Hulchiy O, Strassner H, Gladen BC: Preterm birth in
two urban areas of Ukraine. Obstet Gynecol 2000, 95:752-755.
doi:10.1186/1476-069X-9-56
Cite this article as: Wojtyniak et al.: Association of maternal serum
concentrations of 2,2’, 4,4’5,5’-hexachlorobiphenyl (CB-153) and
1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (p,p’-DDE) levels with
birth weight, gestational age and preterm births in Inuit and European
populations. Environmental Health 2010 9:56.
Submit your next manuscript to BioMed Central
and take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at
www.biomedcentral.com/submit