Environ. Sci. Technol. 2006, 40, 1420-1426
Physical-chemical and Maternal
Determinants of the Accumulation
of Organochlorine Compounds in
Four-Year-Old Children
DANIEL CARRIZO AND
JOAN O. GRIMALT*
Department of Environmental Chemistry, Institute of
Chemical and Environmental Research (IIQAB-CSIC),
Jordi Girona, 18, 08034-Barcelona, Catalonia, Spain
NURIA RIBAS-FITO AND JORDI SUNYER
Respiratory and Environmental Health Research Unit,
Institut Municipal d’Investigació Mèdica (IMIM),
Barcelona, Catalonia, Spain
MATIES TORRENT
AÅ rea de Salut de Menorca, IB-SALUT, Menorca, Spain
A cohort study representing a general population (Minorca
Island, birth year 1997-1998) showed that in utero
transfer of organochlorine compounds (OCs) in children
was strongly correlated with the age of the mother and, in
the case of hexachlorobenzene (HCB), 4,4′-DDE, and 4,4′DDT, with the maternal body mass index. Some of
these correlations remained significant for the serum
concentrations collected in these children at four years.
No significant correlations with length of gestation were
observed. Breastfeeding and age of lactation were strong
determinants of most OC concentrations at four years of
age. At this age, the body burden of these compounds was
higher than at birth irrespective of maternal or formula
feeding, but they accumulated at higher extent in the former
case involving concentration increments in blood that
surpassed the growth dilution effects, with the only exceptions
being pentachlorobenzene (PeCB), HCB, and 4,4′-DDT. 4,4′DDE exhibited the highest increase in association with
breastfeeding, pointing to a specific accumulation pathway
via this mode. Compounds with low Kow values such as
β-hexachlorocyclohexane showed significant accumulation
in four-year-old children but with small differences
between the groups that had been raised on either breast
milk or formula. Compounds having low Koa values such
as HCB showed decreases in concentration and small body
burden variation between birth and four years of age,
which points to their preferential elimination in these initial
periods of infant growth.
the association between exposure to these compounds and
detrimental neurodevelopment effects has not been observed
(6, 7).
These pollutants are incorporated into children in utero
and through diet, mainly through breastfeeding (2, 8-11).
Exposure through breastfeeding is important due to the
widespread occurrence of these compounds, their lipophilicity, and their high stability to chemical degradation, which
favors their accumulation in human fat. These properties
are those characteristic of persistent organic pollutants
(POPs), a group that encompasses the aforementioned
compounds as well as HCB, hexachlorocyclohexanes (HCHs),
and 4,4′-DDT among others, involving up to 12 types of OCs
that have been banned by the Stockholm agreement (12).
Some studies have found that OC blood concentrations
are higher in breastfed than formula fed children even some
years after discontinuation of breastfeeding [e.g., 3.5 years
(13), 4 years (14), or 7 years afterward (9)]. These results
outline the need to improve our knowledge of the real intake
of OCs through maternal and formula feeding within the
first years of child development. Full appraisal of the relevance
of these ingestion modes also requires accurate comparative
estimates with the amounts incorporated in utero. This
information is important in view of the reports of the World
Health Organization (15) and the American Academy of
Pediatrics (16).
In addition, the compound-specific patterns of accumulation through breastfeeding may provide significant
information on the physical-chemical properties that enhance the incorporation of pollutants into children through
this mode. This information may be relevant from the
perspective of implementing new management policies to
evaluate the environmental impact of newly synthesized
chemicals, such as the European Union regulations for
Registration, Evaluation, and Authorization of Chemicals
(REACH) (17).
To gain insight into these questions, a detailed comparison
between OC in cord blood and in sera collected at four years
of age in a cohort of children from Minorca (Balearic Islands,
Mediterranean Sea) was undertaken. The island does not
have factories producing OCs, but DDT had been used for
agriculture in the past. The subjects had therefore been
exposed to baseline POP levels and could thus be taken to
be representative of the regular exposure to these pollutants
in western countries. The cohort recruited all women
presenting for antenatal care over 12 months starting in mid
1997 (18). 482 children were enrolled, and 470 (97.5%)
provided complete outcome data up to the fourth year visit.
Among these, 410 (85%) had OCs measured in cord blood
and 285 (59%) in sera collected at four years.
The results provide insight into the maternal determinants
and compound properties that influence the accumulation
of OCs at birth and at four years of age. They also show the
differences in body burden at this age, after maternal or
formula feeding, in comparison to in utero intake.
Materials and Methods
Introduction
Several studies in children have documented neurodevelopment delays and cognition impairment due to prenatal
and infant exposure to baseline levels of 4,4′-DDE (1, 2) or
polychlorinated biphenyls (PCBs) (1, 3-5). In other cases,
* Corresponding author phone: +34934006122; fax: +34932045904;
e-mail: [email protected].
1420
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 5, 2006
Materials. Standards of tetrabromobenzene (TBB), PeCB,
HCB, R-, β-, γ-, and δ-HCH, PCBs, 4,4′-DDT, and 4,4′-DDE
were purchased from Dr. Ehrenstoffer (Augsburg, Germany).
Analytical grade concentrated sulfuric acid (concentrated
H2SO4), acetonitrile (CH3CN), isooctane, and n-hexane were
purchased from Merck (Darmstadt, Germany).
Extraction. Serum and cord blood samples (0.5 mL) were
introduced into 10 mL centrifuge tubes, and TBB and PCB209
were added as recovery standards. Concentrated H2SO4 (2
10.1021/es0518427 CCC: $33.50
 2006 American Chemical Society
Published on Web 01/27/2006
mL) and n-hexane (3 mL) were then added, and the sample
was mixed in a vortex (ca. 1500 rpm, 30 s) and then centrifuged
(ca. 1500 rpm, 10 min). The supernatant n-hexane layer was
aspirated into a second centrifuge tube using a Pasteur pipet.
Additional n-hexane (2 mL) was added to the first tube, stirred
(vortex ca. 1500 rpm, 30 s), and then centrifuged (ca. 1500
rpm, 10 min). This last step was repeated, yielding 7 mL of
combined n-hexane extracts, to which 2 mL of concentrated
H2SO4 was added. The sample was then mixed (vortex mixer,
ca. 1500 rpm, 90 s), centrifuged as before, and the supernatant
was transferred to a conical bottom, graduated tube. The
combined extracts were then reduced to near dryness under
a gentle stream of nitrogen, at which point PCB142 in
isooctane (10 µL) was added as an injection standard. The
sample was then quantitatively transferred to GC vials using
four 25 µL rinses of isooctane. If an emulsion had been formed
at any stage of the extraction, 10-15 drops of MilliQ water
were added before sample centrifugation.
Instrumental Analysis. A gas chromatograph with electron capture detection (Hewlett-Packard 6890N GC-ECD)
was used to quantify PeCB and HCB, PCB congeners #28,
#52, #101, #118, #138, #153, #180, p,p′-DDT, and p,p′-DDE.
R-, β-, γ-, and δ-HCH were quantified by GC-MS (HP 5973
MSD) in negative chemical ionization mode using ammonia
as reagent gas (1.0 mL/min). In both instruments, samples
were injected (2 µL) in splitless mode onto a 60 m, DB-5
column with a retention gap (both from J&W/Agilent) using
helium as carrier gas (1.5 mL/min). The temperature program
was 90 °C for 2 min, 20 °C/min to 140 °C, 4 °C/min to 200
°C held for 13 min, and finally 4 °C/min to 310 °C held for
10 min.
In both instruments, quantification was performed using
external standards, with the PCB142 injection standard used
to correct for volume. Recoveries of TBB and PCB209 (75115%) were used to correct results. Limits of detection (LOD)
and quantification (LOQ) ranged between 0.02-0.09 and
0.03-0.13 ng/mL, respectively. These values were calculated
from blanks (i.e., the mean of all blanks plus the product of
the standard deviation times three LOD or five LOQ). When
the compound was absent from the blanks, the LOD was
calculated from instrumental data using diluted standards.
This method performed satisfactorily in repeated international intercalibration exercises within the Arctic Monitoring
and Assessment Program (23).
In all cases, when the concentrations were below the LOD,
zero values were introduced. Substitution of these values by
half of the LOQ did not change the results or the significance
of the statistical tests or the correlation analyses discussed
below.
This study was approved by the ethics committee of the
Institut Municipal d’Investigació Mèdica, and all mothers
provided a signed consent form.
Results and Discussion
Characteristics of the Population Studied. Age and body
mass index (BMI) of the participant mothers at delivery are
given in Table 1. The ages represented nearly the whole range
of reproductive activity. BMI encompassed a large spectrum
of cases from underweight (15.3) to obesity (48.5). Time of
gestation is also indicated. Some cases involved short
gestation periods. For children, sex, feeding mode, and length
of lactation are also indicated. 83% of children were breastfed.
Time of lactation ranged from very short (2 months or less)
to very long (more than 1 year). No significant biases between
the group of participants at birth (n ) 410) and four years
later (n ) 285) were observed.
OC Levels. Among the entire cohort, 4,4′-DDE was the
most abundant OC both in cord blood and sera collected at
four years (average 1.6 ng/mL in both cases; Table 2). The
concentrations of this compound were lower than those
TABLE 1. Characteristics of the Study Population
number of
individuals
participants
at birth (at four years)a
sex
male
female
feeding mode
maternal milk
formula milk
time of lactation (weeks)
0.3-10 (0.3-12)a
10-20 (12-21.5)
20-28 (21.5-28)
28-100 (28-96)
time of gestation (weeks)
27-39
39-40
40-44
maternal body mass index
15.3-20.6 (15.3-20.4)a
20.6-22.0 (20.4-22.0)
22.0-24.3 (22.0-24.2)
24.3-48.5 (24.2-48.5)
maternal age
17-26 (17-26)a
26-29 (26-29)
29-32 (29-32)
32-42 (32-41)
%
410 (285)a
202 (136)a
208 (148)
49 (48)a
51 (52)
339 (235)a
71 (49)
83 (83)a
17 (17)
85 (59)a
85 (59)
85 (59)
84 (59)
25 (25)a
25 (25)
25 (25)
25 (25)
106 (72)a
208 (161)
96 (50)
26 (25)a
51 (57)
23 (18)
102 (71)a
103 (72)
103 (71)
102 (71)
25 (25)a
25 (25)
25 (25)
25 (25)
102 (71)a
103 (72)
103 (71)
102 (71)
25 (25)a
25 (25)
25 (25)
25 (25)
a Subset of the same individuals participating in the study at four
years old.
reported in Norway (3.0 ng/mL; 20) and higher than in Canada
(0.4 ng/mL; 10) or Catalonia (0.83 ng/mL; 21). Lower
concentrations of 4,4′-DDT (0.18 and 0.073 ng/mL, respectively) than 4,4′-DDE were observed, likely reflecting that
the whole mixture of DDT metabolites correspond to old
inputs because a substantial amount of the 4,4′-DDT initially
introduced into the environment had already been transformed into 4,4′-DDE (22, 23).
HCB was the second major OC with average values of
0.75 and 0.42 ng/mL in cord blood and four years sera,
respectively (Table 2). These concentrations were lower than
those reported in Norway (1.0 ng/mL; 20), Germany between
1984 and 1985 (2.0 ng/mL; 24), or Catalonia (1.2 ng/mL; 21),
but higher than those found in Germany between 1994 and
1995 (0.61 ng/mL; 24) or Canada (0.04 ng/mL; 10). PeCB was
the OC found in lowest concentration, 0.081 and 0.023 ng/
mL in cord blood and sera collected at four years, respectively.
Total concentrations of the ICES 7 PCB congeners were
0.70 and 1 ng/mL, respectively. The PCB distributions were
dominated by congener PCB153 in both types of samples.
These concentrations were lower than those found in studies
from Norway (3.0 ng/mL; 20), USA (2.5 ng/mL; 25), Faroe
Islands (1.1 ng/mL; 26), and Germany (0.96 or 1.4 ng/mL;
24), but higher than those reported in The Netherlands (0.38
ng/mL; 8), Canada (0.50 ng/mL; 10), or Catalonia (0.36 ng/
mL; 21). Comparison of these data must be done with caution
because different PCB congeners were used for the calculation
of total PCBs in each study. However, the reported figures
may vary by a factor of 2 at the most. In contrast to other
studies, PCB congeners of higher volatility were also considered for quantification in this cohort.
The R-, γ-, and δ-HCH isomers were only found above
quantification limit in less than 5% of total samples. These
compounds were therefore not included in the study. β-HCH
was found at concentrations of 0.21 and 0.28 ng/mL in cord
blood and sera collected at four years, respectively.
Maternal Determinants of OC Concentrations in Children. Cord blood OC concentrations showed significant
VOL. 40, NO. 5, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1421
TABLE 2. Concentrations of the Organochlorine Compounds in Cord Blood and Blood at Four Years in the Study Population, and
Their Octanol-Water and Octanol-Air Partition Coefficients
cord blood
(n ) 410)
e
compounds
(ng/mL)
PeCBa
HCBb
β-HCHc
PCB-28
PCB-52
PCB-101
4,4′-DDE
PCB-118
PCB-153
4,4′-DDT
PCB-138
PCB-180
sumPCBd
blood at four years
(n ) 285)
average
standard
deviation
average
standard
deviation
0.081
0.75
0.21
0.014
0.021
0.032
1.6
0.078
0.21
0.18
0.17
0.20
0.68
0.27
0.79
0.42
0.070
0.076
0.097
2.0
0.084
0.24
0.27
0.13
0.36
0.71
0.023
0.42
0.28
0.024
0.036
0.088
1.6
0.10
0.35
0.073
0.24
0.20
1.0
0.085
0.43
0.43
0.33
0.30
0.14
3.2
0.12
0.67
0.12
0.53
0.47
2.5
a Pentachlorobenzene. b Hexachlorobenzene. c β-Hexachlorocyclohexane.
Calculated from refs 32, 33.
d
log(Kow)
log(Koa)
25 °C (36 °C)e
25 °C (36 °C)e
4.92 (4.78)
5.48 (5.38)
3.8 (3.7)
5.71 (5.54)
5.83 (5.60)
6.35 (6.30)
5.95 (5.90)
6.69 (6.55)
6.75 (6.65)
6.16 (6.10)
6.69 (6.59)
7.19 (7.15)
6.46 (6.02)
7.35 (6.95)
8.88 (8.26)
8.06 (7.59)
8.14 (7.60)
9.00 (8.49)
9.33 (8.71)
9.66 (9.09)
9.62 (9.06)
9.99 (9.43)
9.71 (9.16)
10.23 (9.70)
Sum of the seven individual PCB congeners analyzed individually.
TABLE 3. Correlations of Levels of Organochlorine Compounds in Children with Length of Pregnancy and Age and Body Mass
Index of the Mother at Deliverya
age
cord blood
(n ) 410)
body mass index
sera at four years
(n ) 285)
cord blood
(n ) 410)
period of pregnancy
sera at four years
(n ) 285)
cord blood
(n ) 410)
compounds
r coefficient
t value
r coefficient
t value
r coefficient
t value
r coefficient
t value
r coefficient
t value
PeCB
HCB
β-HCH
PCB-28
PCB-52
PCB-101
4,4′-DDE
PCB-118
PCB-153
4,4′-DDT
PCB-138
PCB-180
sumPCBs
0.024
0.20
0.19
0.010
0.033
0.010
0.021
0.099
0.15
0.12
0.19
0.057
0.077
0.48
4.2****
3.9****
0.19
0.67
0.19
4.4****
2.0*
3.0***
2.5***
3.8****
1.2
1.7*
0.005
0.17
0.075
0.029
0.033
0.016
0.048
0.032
0.035
0.12
0.025
0.043
0.016
0.089
2.8**
1.3
0.49
0.56
0.27
0.80
0.54
0.58
2.1*
0.41
0.73
0.27
0.041
0.23
0.077
0.005
0.05
0.05
0.17
0.011
0.043
0.11
0.035
0.040
0.041
0.82
4.8****
1.5
0.092
1.0
1.1
3.5***
0.23
0.88
2.3*
0.71
0.80
0.88
0.082
0.14
0.065
0.087
0.085
0.093
0.086
0.069
0.063
0.034
0.067
0.071
0.076
1.4
2.4*
1.1
1.5
1.4
1.6
1.5
1.2
1.1
0.58
1.1
1.2
1.3
0.016
0.042
0.035
0.050
0.025
0.08
0.024
0.005
0.033
0.074
0.022
0.089
0.088
0.32
0.86
0.71
1.2
0.50
1.6
0.49
0.094
0.66
1.5
0.44
1.8
1.9
a
*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.
correlations with the age of the mother at delivery for HCB,
β-HCH, 4,4′-DDE, 4,4′-DDT, PCB118, PCB153, PCB138, and
total PCBs (Table 3). According to these results, older mothers
transferred higher OC concentrations into newborns. The
compounds exhibiting these correlations were those found
in higher concentration in cord blood (Table 2). In fact, with
the exception of PCB180, all measured individual OC at
average concentrations higher than 0.1 ng/mL exhibited
positive correlations with very high significance (p < 0.001).
Maternal age has also been described in other studies as a
predictor of cord blood levels of PCBs and HCB (27), 4,4′DDE (21) or 4,4′-DDE, PCBs, and HCB (10). The concentrations in sera collected at four years only showed significant
correlation with age of the mother for HCB (p < 0.01) and
4,4′-DDT (p < 0.05). The incorporation of new OC amounts
through diet (e.g., breastfeeding) probably decreased the
relevance of the initial in utero input significantly except for
the case of the two aforementioned compounds.
OC concentrations in cord blood showed significant
correlations with the BMI of the mother at delivery for HCB,
4,4′-DDE, and 4,4′-DDT (Table 3). Higher BMI corresponded
to higher OC concentrations in cord blood. No significant
association between the concentrations of these compounds
and cord blood lipids was observed. For HCB and 4,4′-DDE,
1422
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 5, 2006
the degree of significance was very high (p < 0.0001 and p
< 0.001, respectively). These two compounds were those
present in the highest average concentration in the population of newborns (Table 2). The concentrations in sera of
four year old children only showed significant correlation
with maternal BMI for HCB (p < 0.05) (Table 3). The same
significant correlations for cord blood and sera collected at
four years were observed when the OC concentrations were
compared to the weight of the mother.
Older gestational age has also been related to higher OC
levels in cord blood (10, 27). However, this association was
not found in the present study (p > 0.09; Table 3) or in a
previous study of a cohort from Ribera d’Ebre (Catalonia)
(28).
The correlations of the cord blood OC concentrations
with maternal age and BMI are consistent with higher OC
accumulation in mothers of higher age and BMI (28), whereby
the OCs from the mothers are ultimately transferred to child
in utero. To the best of our knowledge, this is the first report
of a direct relationship between maternal BMI and the
concentration of some OCs in children at birth and at four
years old.
Influence of Milk Feeding. The average concentrations
of HCB, 4,4′-DDE, 4,4′-DDT, PCB congeners #153, #138, and
FIGURE 1. Average concentrations of organochlorine compounds in cord blood and sera collected from four-year-old children who had
been raised on either breast milk or formula. The intervals represent standard error (p < 0.05) and show the values corresponding to the
whole datasets and the subset of common samples for newborns and four-year-old children, respectively. Asterisks indicate significant
differences between the means of children raised on either breast milk or formula (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).
#180, and total PCBs in sera collected at four years exhibited
significantly higher values in breastfed than artificially fed
children (Figure 1). The degree of significance of the
differences was very high (p < 0.0001) for most of these
compounds. Calculation of the average values over the whole
dataset of samples collected at four years (n ) 285) or the
subset of children providing samples at four years and cord
blood (n ) 244) showed nearly the same degree of significance
(Figure 1). The internal consistency of the database was then
confirmed by calculation of the same means using the cord
blood data. As expected, no significant difference (p > 0.05)
was observed between maternally and formula fed children
(in either the databases of n ) 285 or n ) 244) (Figure 1). The
results of Figure 1 demonstrate that breastfeeding is very
significant for the concentrations of OC in four-year-old
children (i.e., children who stopped breastfeeding 2.3-3.5
years before being tested).
In agreement with this observation, the period of lactation
was also correlated with the concentrations of HCB, β-HCH,
4,4′-DDE, PCB118, PCB153, 4,4′-DDT, PCB138, PCB180, and
total PCBs accumulated in four-year-old breastfed children
(Table 4). In all cases, longer lactation corresponded to higher
concentrations in sera. Examination of the same correlations
using cord blood data did not show any significant correlation
(Table 4). Again, this result was expected, but, as in the
previous case, it was calculated to check for the internal
consistency of the database.
The studies showing that breast milk is an important
source for the incorporation of OCs into children even several
TABLE 4. Correlations of Periods of Lactation with Levels of
Organochlorine Compounds at Four Years of Age
cord blood
(n ) 339)a
blood at four years
(n ) 235)a
compounds
r coefficient
t value
r coefficient
t value
PeCB
HCB
β-HCH
PCB-28
PCB-52
PCB-101
4,4′-DDE
PCB-118
PCB-153
4,4′-DDT
PCB-138
PCB-180
sumPCBs
0.09
0.05
0.007
0.043
0.099
0.006
0.051
0.028
0.11
0.11
0.071
0.072
0.087
1.7
0.91
0.13
0.78
1.8
0.11
0.94
0.52
1.9
1.9
1.3
1.3
1.6
0.003
0.41
0.26
0.014
0.047
0.026
0.24
0.14
0.19
0.28
0.17
0.15
0.13
0.044
6.9***
4.1***
0.22
0.72
0.40
3.7***
2.2*
3.0**
4.5***
2.6**
2.3*
2.0*
a Calculated over the population of breastfed children (the analysis
using cord blood was also performed for reference).*p < 0.05, **p <
0.01, ***p < 0.001, ****p < 0.0001.
years after breastfeeding had stopped are based on correlations between duration of lactation and serum OC concentrations. Mixtures of 4,4′-DDE (9), 4,4′-DDT (14), β-HCH (9),
HCB (9), and PCB having diverse congeners with more than
five chlorine substitutents (9, 13, 14) have been found in
higher concentration in children fed with maternal milk for
VOL. 40, NO. 5, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1423
FIGURE 2. Average concentration and body burden differences between cord blood of newborns and sera of four-year-old children raised
on either breast milk or formula. The intervals represent standard error (p < 0.05) and show the values corresponding to the whole datasets
and the subset of common samples for newborns and four-year-old children, respectively. The body burden differences averaged the
individual body burden calculations of children from which cord blood and sera collected at four years were available.
long periods. The results observed in the population of
Minorca are in agreement with these previous results. In this
case, the nursing period encompassed a very wide time range
(0.3-100 weeks), and nearly all OCs showed significant
correlation with this determinant (Table 4). Only PeCB and
the PCB congeners having three (e.g., #28), four (e.g., #52),
and some having five (e.g., #101) chlorine atoms were not
correlated to lactation time.
Further understanding of the relevance of breastfeeding
to overall OC intake can be obtained by comparing the
average values between maternally and formula fed children.
In the Minorca cohort, the average concentrations of HCB,
4,4′-DDE, 4,4′-DDT, PCB congeners #153, #138, #180, and
total PCB were significantly higher in subjects that had been
breastfed in infancy (Figure 1). These results were in close
agreement with the aforementioned correlations with duration of breastfeeding as well as with those of previous studies.
However, β-HCH and PCB118 whose concentrations were
observed to be directly proportional to duration of breastfeeding (Table 4) had average values that were not significantly different between the two feeding mode groups (p >
0.05) (Figure 1). This contrast indicates that, despite higher
incorporation of these compounds with milk ingestion, their
background intake due to formula feeding or in utero
exposure was high enough to diminish the significance of
breastfeeding (e.g., average β-HCH concentrations in maternal and formula feeders were 0.30 and 0.20 ng/mL,
respectively). This result shows that, independently of
correlations with duration of breastfeeding, statistical comparisons of average OC concentrations for each feeding mode
1424
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 5, 2006
are needed to determine the real contribution of breastfeeding to OC intake in children.
Changes in OC Concentrations. In breastfed children,
the average concentration differences between sera collected
at four years and cord blood showed an increase in the
concentrations of all OC, with the only exceptions being PeCB,
HCB, and 4,4′-DDT (Figure 2). No change was observed for
PCB180. The pattern of these differences was entirely distinct
in formula fed children. In this case, lower levels were
observed for nearly all OCs. Similar results were found when
the calculations were performed over the whole population
(n ) 410 and 285 for cord blood and sera at four years,
respectively) or only over the subjects who provided both
cord blood and sera at four years (n ) 244).
The magnitudes of change were different for maternally
and formula fed groups. The increments associated with the
former were much smaller (0.5 ng/mL at the most) than the
decreases observed for the latter (up to 1.5 ng/mL) (Figure
2). As expected, dilution resulting from growth of the children
tended to reinforce the decreases and counterbalance the
increases. In this cohort, the average growth involved changes
from ca. 3.2 kg at birth to ca. 16.2 kg at four years of age
corresponding to approximate blood volumes of 0.24 and
1.2 L, respectively.
The distinct concentration values observed for PeCB, HCB,
and 4,4′-DDT in breastfeeders can be related to their low
log(Koa) values (<7 at 36.5 °C), which were significantly lower
than the constants of the other compounds (Table 2). The
significance of the Koa constants was more evident when
considering that β-HCH, which has the lowest Kow of the
whole group of compounds but has a log(Koa) > 8, exhibited
the same behavior as the other OCs studied. The low extent
of accumulation of PeCB and HCB indicates that the more
volatile compounds are less retained. Exhalation is a possible
efficient route of excretion in these early periods of infant
growth (29). 4,4′-DDT does not exhibit Kow and Koa constants
that are significantly different from the other OCs that show
increments at four years. The distinct behavior of this
compound can be explained by transformation into 4,4′DDE, a phenomenon commonly observed in mammals (23).
Changes in OC Body Burden. The weight of each child
at birth and at four years of age was compiled for an estimate
of the blood volume (assuming 75 mL/kg; 30), which in
combination with the aforementioned OC concentrations
allowed estimation of differences in body burden among
participants. As shown in Figure 2, the average body burdens
of all OC increased in both breast milk and formula groups.
However, significant contrasts were observed between the
two feeding mode groups.
In the breastfeeding group, the major increase was
observed for 4,4′-DDE (1.8 µg) and the second for total PCB
(1.3 µg). The PCB increase was dominated by increments of
the more chlorinated congeners such as #153 (0.5 µg), #138
(0.3 µg), and #180 (0.2 µg), whereas the less chlorinated
congeners (e.g., #28 and #52), exhibited small increases (<0.1
µg). These results were consistent with a group of human
milk samples analyzed in the context of this work for reference
and general studies showing that the more chlorinated
congeners are those predominant in breast milk (31). HCB
and β-HCH increased by 0.5 and 0.3 µg, respectively. The
high levels of 4,4′-DDE incorporation through breastfeeding
were consistent with previous observations for other cohorts
(11). The high accumulation of this compound could not be
explained by specific OC physical-chemical properties. Thus,
it may have reflected a specific metabolic mechanism of
incorporation into breast milk, which would enhance its
intake in relation to other OCs.
In artificial feeding, the major increment was observed
for total PCBs (0.6 µg) involving higher increases of the
congeners with five and six chlorine atoms, and lower
increases of the others. The increase in 4,4′-DDE was very
little (0.1 µg) as compared that in breastfeeding. The small
increases of PeCB and HCB, 0.06 and 0.04 µg, respectively,
are consistent with their low Koa as discussed above. In
contrast, β-HCH exhibited an increment of 0.2 µg and was
the compound for which changes were most similar in both
feeding modes. This OC can be differentiated from the others
included in the study by its low log(Kow) values (3.8 at 36.5
°C, Table 2). Thus, this OC was the least hydrophobic of the
whole group, suggesting that its incorporation into children
via breastmilk is relatively low. However, after intake, this
compound accumulates in a fashion similar to the other OC
that have log(Koa) > 8, which reinforces the aforementioned
results and indicates a preferential elimination of those
compounds with low Koa.
Acknowledgments
We thank L. Font and M. Gari for their help in sample analysis.
M. C. Alvaro is thanked for her useful advice on the
interpretation of results. This research was supported by the
Instituto de Salud Carlos III, Red de Grupos INMA (G03/
176), and project PI041666.
Literature Cited
(1) Rogan, W. J.; Gladen, B. C.; McKinney, J. D.; Carreras, N.; Hardy,
P.; et al. Neonatal effects on transplacental exposure to PCBs
and DDE. J. Pediatr. 1986, 109, 335-341.
(2) Ribas-Fito, N.; Cardo, E.; Sala, M.; de Muga, E.; Mazon, C.; et
al. Breastfeeding, exposure to organochlorine compounds, and
neurodevelopment in infants. Pediatrics 2003, 111, e580-e585.
(3) Darvill, T.; Lonky, E.; Reihman, J.; Stewart, P.; Pagano, J. Prenatal
exposure to PCBs and infant performance on the Fagan test of
infant intelligence. Neurotoxicology 2000, 21, 1029-1038.
(4) Jacobson, J. L.; Jacobson, S. W. Intellectual impairment in
children exposed to polychlorinated biphenyls in utero. N. Engl.
J. Med. 1996, 335, 783-789.
(5) Walkowiak, J.; Wiener, J. A.; Fastabend, A.; Heinzow, B.; Kraemer,
U.; et al. Environmental exposure to polychlorinated biphenyls
(PCBs) and quality of the home environment: effects on
psychodevelopment in early childhood. Lancet 2001, 358, 16021607.
(6) Grandjean, P.; Weihe, P.; Burse, V. W.; Needham, L. L.; StorrHansen, E.; et al. Neurobehavioral deficits associated with PCB
in 7-year-old children prenatally exposed to seafood neurotoxicants. Neurotoxicol. Teratol. 2001, 23, 305-317.
(7) Gray, K. A.; Klebanoff, M. A.; Brock, J. W.; Zhou, H.; Darden, R.;
et al. In utero exposure to background levels of polychlorinated
biphenyls and cognitive functioning among school-age children.
Am. J. Epidemiol. 2005, 162, 17-26.
(8) Huisman, M.; Koopman-Esseboom, C.; Fidler, V.; Hadders-Algra,
M.; van der Paauw, C. G.; et al. Perinatal exposure to polychlorinated biphenyls and dioxins and its effect on neonatal
neurological development. Early Hum. Dev. 1995, 41, 111-127.
(9) Karmaus, W.; deKoning, E. P.; Kruse, H.; Witten, J.; Osius, N.
Early childhood determinants of organochlorine concentrations
in school-aged children. Pediatr. Res. 2001, 50, 331-336.
(10) Rhainds, M.; Levallois, P.; Ayotte, P. Lead, mercury, and
organochlorine compound levels in cord blood in Québec,
Canada. Arch. Environ. Health 1999, 54, 40-47.
(11) Ribas-Fito, N.; Grimalt, J. O.; Marco, E.; Sala, M.; Mazon, C.; et
al. Breastfeeding and concentrations of HCB and p,p′-DDE at
the age of 1 year. Environ. Res. 2005, 98, 8-13.
(12) http://www.pops.int, 2005.
(13) Lanting, C. I.; Patandin, S.; Fidler, V.; Weisglas-Kuperus, N.;
Sauer, P. J. J.; et al. Neurological condition in 42-month-old
children in relation to pre- and postnatal exposure to polychlorinated biphenyls and dioxins. Early Hum. Dev. 1998, 50,
283-292.
(14) Jacobson, J. L.; Humphrey, H. E. B.; Jacobson, S. W.; Schantz,
S. L.; Mullin, M. D.; et al. Determinants of polychlorinated
biphenyls (PCBs), polybrominated biphenyls (PBBs), and
dichlorodiphenyl trichloroethane (DDT) levels in the sera of
young children. Am. J. Public Health 1989, 79, 1401-1404.
(15) Brouwer, A.; Ahlborg, U. G.; van Leeuwen, F. X.; Feeley, M. M.
Report of the WHO working group on the assessment of health
risks for human infants from exposure to PCDDs, PCDFs and
PCBs. Chemosphere 1998, 37, 1627-1643.
(16) Committee on Environmental Health. Handbook of Pediatric
Environmental Health. Elk Grove Village, IL. Am. Acad. Pediatr.
1999, 155-162.
(17) http://europa.eu.int/comm/environment/chemicals/, 2005.
(18) Polk, S.; Sunyer, J.; Munoz-Ortiz, L.; Barnes, M.; Torrent, M.;
Figueroa, C.; et al. A prospective study of Fel d1 and Der p1
exposure in infancy and childhood wheezing. Am. J. Respir.
Crit. Care Med. 2004, 170, 273-8.
(19) http://www.amap.no.
(20) Skaare, J. U.; Tuveng, J. M.; Sande, H. A. Organochlorine
pesticides and polychlorinated biphenyls in maternal adipose
tissue, blood, milk, and cord blood from mothers and their
infants living in Norway. Arch. Environ. Contam. Toxicol. 1988,
17, 55-63.
(21) Sala, M.; Ribas-Fito, N.; Cardo, E.; deMuga, M. E.; Marco, E.; et
al. Levels of hexachlorobenzene and other organochlorine
compounds in cord blood: exposure across placenta. Chemosphere 2001, 43, 895-901.
(22) Wedemeyer, G. Dechlorination of 1,1,1-trichloro-2,2-bis(pchlorophenyl)ethane by Aerobacter aerogenes. I Metabolic
products. Appl. Microbiol. 1967, 15, 569-574.
(23) Aguilar, A. Relationship of DDE/sigma DDT in marine mammals
to the chronology of DDT input into the ecosystem. Can. J. Fish.
Aquat. Sci. 1984, 41, 840-844.
(24) Lackmann, G. M.; Goën, T.; Töllner, U.; Schaller, K. H.; Angerer,
J. PCBs and HCB in serum of full-term German neonates. Lancet
1996, 348, 1035.
(25) Schwartz, P. M.; Jacobson, S. W.; Fein, G.; Jacobson, J. L.; Price,
H. A. Lake Michigan fish consumption as a source of polychlorinated bipenyls in human cord serum, maternal serum,
and milk. Am. J. Public Health 1983, 73, 283-296.
(26) Grandjean, P.; Weihe, P.; White, R. F.; Debes, F.; Araki, S.; et al.
Cognitive deficit in 7-year-old children with prenatal exposure
to methylmercury. Neurotoxicol. Teratol. 1997, 19, 417-428.
VOL. 40, NO. 5, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1425
(27) Lackmann, G. M.; Angerer, J.; Salzberger, U.; Töllner, U.; et al.
Influence of maternal age and duration of pregnancy on serum
concentrations of polyclorinated biphenyls and hexachlorobenzene in full-term neonates. Biol. Neonate 1999, 76, 214219.
(28) Schade, G.; Heinzow, B. Organochlorine pesticides and polychlorinated biphenyls in human milk of mothers living in
northern Germany: current extent of contamination, time trend
from 1986 to 1997 and factors that influence the levels of
contamination. Sci. Total Environ. 1998, 215, 31-39.
(29) Daston, G.; Faustman, E.; Ginsberg, G.; Fenner-Crisp, P.; Olin,
S.; Sonawane, B.; Bruckner, J.; Breslin, W. A framework for
assessing risks to children from exposure to environmental
agents. Environ. Health Perspect. 2004, 112, 238-256.
(30) http://www.drgreene.com/21_1616.html.
(31) Schecter, A.; Furst, P.; Krüger, C.; Meemker, H. A.; Groebel, W.;
Constable, J. D. Levels of polychlorinated dibenzofurans,
1426
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 5, 2006
dibenzodioxins, PCBs, DDT and DDE, hexachlorobenzene
(HCB), dieldrin, hexachlorocyclohexane and oxychlordane in
human breast milk from the United States, Thailand, Vietnam,
and Germany. Chemosphere 1989, 18, 445-454.
(32) Beyer, A.; Wania, F.; Gouin, T.; Mackay, D.; Matthies, M. Selecting
internally consistent physicochemical properties of organic
compounds. Environ. Toxicol. Chem. 2002, 21, 941-953.
(33) Schoeib, M.; Harner, T. Using measured octanol-air partition
coefficients to explain environmental partitioning of organochlorine pesticides. Environ. Toxicol. Chem. 2002, 21, 984990.
Received for review September 17, 2005. Revised manuscript
received December 22, 2005. Accepted December 27, 2005.
ES0518427