Infections and atopic disorders in childhood and organochlorine exposure

Infections and atopic disorders in childhood and organochlorine exposure

Wilfried Karmaus

IN 1995, we conducted a large-scale environmental epidemiological study in the south of the Federal State of Hessen, Germany. Children were recruited from 3 regions. One region, located in the Rhine Valley and surrounded by low mountains on both sides, is approximately 30 km from an industrial waste incinerator and other industries (e.g., chemical plants). Several municipalities lie in the area affected environmentally by these industrial sites. The region is also used intensively for the production of vegetables. Another region, also industrial and agricultural, is 20 km north of the incinerator area. The third region is located in a low-mountain area (i.e., approximately 400 m above sea level). Results on polychlorinated biphenyls (PCBs), thyroid hormones, chromium, and lymphocytes have been published elsewhere. (1-3)

The current study was initially prompted by a recent report on the susceptibility to infection, particularly otitis media, in Inuit infants exposed to organochlorine compounds (OCs). (4) In this investigation, however, investigators were unable to isolate the specific OC to which the infectious diseases were attributable. Second, the current work was motivated by a publication that reported higher immunoglobulin (Ig)-E levels in cord blood following in utero dioxin exposure. (5) Increased IgE is a hallmark of atopic (allergic) diseases, such as asthma. (6,7) In addition, manifestations of infectious diseases and of allergic reaction are seen as a dichotomous pattern of response. Immunologically, the T-helper cell-subset, Th1, promotes immune response (e.g., against infections). Th2 cells stimulate humoral immunity and IgE production.

Thus, our concern was focused on whether the 2 outcomes–infections and atopic manifestations–were associated with OCs in our sample of children. We focused on indicators of infections: parental information about otitis media, pneumonia, and whooping cough. As indicators of atopy, we included asthma, asthma symptoms, and IgE serum concentration in a cross-sectional analysis.

Method

Study population. After obtaining permits from the Data Protection Agency of the State of Hamburg (Germany), from the Ministry of Cultural Affairs of Hessen (Germany), and from the local school committees, we asked the parents of 1,091 second-grade school children in 18 townships to participate in our study. We obtained informed consent from all participating parents, in accordance with requirements of the Ethical Committee of the Board of Physicians and the Data Protection Agency of the State of Hamburg. We asked parents to allow their child to participate in phlebotomy only when passive smoking in the private household did not exceed 10 cigarettes/day during the previous 12 mo.

Questionnaires. We used 4 self-administered questionnaires in the survey. Each of the questionnaires solicited information about the living conditions and nutrition of the families, 1 each for the mother and the father, and 2 for the data of the child. We asked if a doctor had diagnosed the child with otitis media (e.g., never, once, twice); whooping cough (pertussis–never vs. once); and pneumonia (never vs. once). In addition, we inquired if the child had ever had asthma (question of the International Study of Asthma and Allergy in Children), (8) and if the child had ever experienced (a) wheezing or whistling, (b) cough or shortness of breath early in the morning or at night, (c) cough after exercise or when exposed to cold air or fog, or (d) attacks of dyspnea. We assessed environmental tobacco smoke (ETS) as smoking in the child’s home during the previous 12 mo (i.e., no cigarettes, 1-10 cigarettes/day, 11-20 cigarettes/day, 21-30 cigarettes/day, and more than 30 cigarettes/day). We asked if the child had been breast-fed and the duration with and without supplemental feeding. We used the child’s vaccination card to determine the number of pertussis vaccinations given.

Immunoglobulin E. One parent accompanied the child during the medical examination. Physicians used the Vacutainer[R] System to withdraw 25 ml of blood after which it was separated into different aliquots. Measurements of IgE in serum were made at the Medical, Alimentary and Veterinary Institute for Research Middle Hessen, Division of Human Medicine, Dillenburg, Germany, with a radioimmunoassay (CAP, Pharmacia [Dillenburg, Germany]). The results were provided in kU/l serum.

Organochlorines in blood. Dichlorodiphenyldichloroethene (DDE), hexachlorobenzene (HCB), hexachlorocyclohexane (HCH), and 8 PCB congeners (i.e., 101, 118, 138, 153, 170, 180, 183, and 187) were analyzed at the Institute of Toxicology, University of Kiel, Germany. The OCs were determined in 5-ml samples of whole blood subjected to high-resolution gaschromatography (Model 3400, Varian Co. [Darmstadt, Germany]) with a 63Ni electron capture detector. The detection limit (DL [i.e., 2 times the signal/low-noise ratio]) was 0.02 [micro]g/l for [beta]- and [gamma]-HCH, DDE, and each PCB congener; the DL was 0.01 [micro]g/l for HCB and [alpha]-HCH. For extraction and clean-up procedures, florisil (9 gm) and n-hexane were used, deactivated with 3% water and placed in a chromatography column for elution; the column was 22 mm in diameter and 48 mm in length for elution. The capillary column was 0.25 mm in diameter and 30 mm long, and nitrogen was the carrier gas. The congeners were determined by retention times on the chromatograms and were identified by comparison with known standards. In addition, reliability was tested with gas chromatography/mass spectrometry.

Statistical method. All statistical analyses were performed with Statistical Analysis System (version 8) software. (9) We calculated the sum of the sample values of the PCB congeners (sum of 7 congeners; 1 congener [PCB101] was not detected). For descriptive purposes, the sample values of the OCs were substituted with one-half of the DL for values below the DL. Inasmuch as the distributions of the OCs were not normal, the geometric mean (GM), median, and 5th and 95th percentiles are provided. We used cross-tabulations and rank correlation coefficients to investigate bivariate association. We used statistical means to group the OCs into quartiles (PROC RANK). All observations below the DL were part of the lowest level group (reference). Therefore, the grouping did not represent toxicological considerations, and no recommended thresholds are known. For [gamma]-HCH, only 3 groups could be defined (tertiles). Later, the distribution of the OC was further condensed into 2 groups (two quartiles combined). Each child’s age and otitis media were categorized (7 yr, 8 yr, [greater than or equal to] 9 yr; ever vs. never in each age group).

We used logistic regression models to analyze risk factors for the prevalence of otitis media, pneumonia, pertussis, and bronchial asthma, and for having an IgE concentration of [greater than or equal to] 200 kU/l serum. Variables of interest were HCB; DDE; PCBs; and [alpha]-, [beta]-, and [gamma]-HCH. Odds ratios (ORs) and their 95% confidence intervals (CIs) were estimated. The distribution of IgE (after log-transformation to achieve a normal distribution) was analyzed with linear regression. GMs of IgE for the different levels of the exposures and p values were estimated.

The effects were adjusted for gender, age, breast-feeding, and ETS as potential confounders. For PCB, DDE, HCB, and HCH, we used dummy variables to indicate each quarter of the respective distribution to check their linear relationship. To check collinearity, we analyzed the tolerance (PROC REG) of the predictors.

Results

The proportion of participation was 61.5% (671 of 1,091). We obtained blood samples from 350 children and were able to conduct complete OC analyses on 343 samples; in addition, we were able to determine IgE in 340 samples. Overall information (i.e., on the questionnaires and about OCs) could be analyzed statistically in 343 children. Fewer girls than boys participated in phlebotomy, and 96% of the children were 7 yr or 8 yr of age (Table 1). Given our selection, the prevalence of cigarette smoking in the child’s home was also lower in the group with phlebotomy than in the total group (Table 1). There was no statistically significant difference with respect to doctors’ diagnoses, asthma, and pertussis vaccination. About 89% of the sample had no vaccination against pertussis. Among the 340 IgE measurements, 15.6% were above a reference of 200 kU/l.

For [alpha]-HCH, 95% of the observations were below the DL; therefore, we excluded this OC from the current statistical analysis. For [gamma]-HCH, 27.7% of the observations were below the DL, 0.58% of the [beta]-HCH observations were below the DL, and none of the DDE and HCB observations were below the DL. At least 1 PCB congener was detected in each sample. With the exception of [gamma]-HCH and HCB, all OCs were correlated (Tables 2 and 3). The highest correlations were found between HCB and the sum of the 7 PCB congeners. These correlations caused numerical problems in the logistic regressions (multi-collinearity).

The results of logistic regression (i.e., explaining the occurrence of infections and atopic manifestations) were sensitive for the inclusion of all 5 OCs as predictors. Some ORs and their 95% CIs had extreme values. Elimination of different single OCs changed the effects; therefore, we simplified the models. First, we excluded [beta]- and [gamma]-HCH, inasmuch as HCHs showed only a few unstable effects in earlier models containing all OCs. We also excluded passive smoking (i.e., during the past 12 mo) from the logistic regression model because its effect was minimal. Second, we included only variables for which effects were more stable–in this case DDE. Third, we added other predictors (e.g., the sum of PCB and HCB) to the model after we stratified for DDE (< 0.3 [micro]g/l vs. [greater than or equal to] 0.3 [micro]g/l), which avoided the earlier-cited numerical problems. This approach enabled us to estimate the combined effects (interaction) of DDE and HCB or DDE and PCB, respectively.

About 20% of the children had experienced otitis media once, and about 40% had it twice (Table 1). The prevalence (yes vs. no) was lower in children with higher blood DDE concentrations (0.3 [micro]g/l, Table 4). In the group with higher DDE values, PCBs and HCB increased the prevalence of otitis media (OR = 3.7 and OR = 2.38, respectively, Table 5). A similar pattern was detected for a history of whooping cough (Tables 4 and 5), but not for the association between pneumonia and PCB. The occurrences of otitis media, pneumonia, and pertussis in the same child were not associated significantly.

The pattern shown for infectious diseases and DDE was the inverse of that shown for DDE and allergic (atopic) disorders. The prevalence of asthma was highest in the upper half of the DDE distribution (Table 4). The frequency of the disease increased with the quartiles of the DDE blood levels (i.e., 3.7%, 3.3%, 4.8%, and 8.1% [data not shown]). The OR for asthma, under mutual control of other predictors, was 3.71 times higher for children in the upper half of the DDE distribution. The 95% CI was wide because of the rarity of disease (Table 5). For DDE, we also identified an increased relative risk for an IgE serum concentration that exceeded 200 kU/l (OR = 2.28; 95% CI = 1.24, 4.31 [Table 5]). In addition, the results of the linear regression of the IgE concentration (including all initial predictors) revealed that, in the lowest DDE quartile, the GM of IgE was 21.1 kU/l, and in the lowest DDE quartile, the GM of IgE was 62.1 kU/l, following a dose-response pattern (Table 6).

The two markers of asthma were associated. Of those children with asthma (n = 17), 10 (58.8%) had an IgE level that exceeded 200 kU/l, but only 13.5% (44/326) without asthma had such an IgE level. Sixteen of the 17 children with a history of asthma had parents who had also reported at least 1 of 4 asthma-like symptoms within the past 12 mo. In the absence of a history of asthma, the symptoms occurred in only 20.6% of the children.

Breast-feeding did not affect incidence of infectious diseases, but it was protective against asthma (OR = 0.27 [Table 5]). Gender affected only atopic manifestations (e.g., girls had 0.1 7 times less asthma than boys). Passive smoking during the 12 months that preceded phlebotomy increased the IgE concentration (Table 6), but was not statistically significant.

Discussion

In 343 school children aged 7-10 yr, we analyzed the relationship between OCs and 3 infectious diseases and 3 atopic manifestations. The work was stimulated by recent publications. (4,5) We observed weak associations for the sum of PCBs, HCB, and infection. However, DDE blood concentration was a consistent, strong risk factor for asthma and IgE.

In this study, we selected a subgroup for blood analyses that had a sufficiently low ETS exposure in their homes to reduce the effects of ETS. Parents of 501 children reported an ETS exposure of 10 or fewer cigarettes per day in the children’s home in the preceding 12 mo. Of this group, 317 participated in the phlebotomy (63%). In addition, parents of 26 of 162 children with a higher ETS exposure (15.4%) managed to have their children included. Environmental tobacco smoke, however, was not related significantly to any of the 3 OCs. Age, gender, and disorders were not substantially different between the total sample and the subgroup (n = 343) for OCs.

When parents provided information about their child’s health and living conditions, they did not know the individual results of blood analyses. Therefore, the information about OCs or IgEs could not have been biased by the selection of the subgroup or by questionnaire or interview data.

In this study, we concentrated on asthma and on 3 infections that occur in childhood (i.e., otitis media, pneumonia, and whooping cough) because these constituted events that required a doctor’s visit and were likely to be recalled. The International Study of Asthma and Allergies in Childhood question of “ever asthma” (8) represented the lifetime prevalence; however, it was closely associated with asthma-like symptoms reported for the most-recent last-12-mo period. IgE serum levels represented the current status. Ever asthma and IgE were also associated strongly, thus indicating that these markers measured a common characteristic of an atopic reaction.

The cross-sectional character of the analysis limited conclusions about the time order. Nevertheless, it seemed reasonable that infections or atopic manifestations followed OC exposure–and not vice versa. In the literature, there was no evidence that atopic disorders could change the exposure or the body distribution of OCs. We assumed that the concentration of the OCs did not vary substantially in childhood, but it represented exposure during the first years of life. For example, the concentration of PCBs in this group of children of about 8 yr of age was also predicted by breast-feeding. (1) The assumption of the stability of the OCs was further supported by a 2-yr follow-up of the children and by determination of DDE, HCB, and PCBs. Spearman’s rank correlation between the 2 successive measurements in the same children was as follows: DDE–r = .86, n = 274, p = .0001; HCB–r= .74, n = 274, p = .0001; and sum of the 7 PCBs–r= .82, n = 274, p = .0001.

For some PCBs, analogous analyses were available for the years 1993 and 1994 from the study of the Federal State of Baden-Wurttemberg, (10) which is adjacent to the current study region. In our study, the GM was slightly higher for PCB138 and PCB153 (2) than in the aforementioned study. The GM showed a comparable concentration for PCB180. However, this similarity might have been biased by different laboratory procedures. The whole blood concentration of HCB was higher in our samples from South Hessen, compared with those from 3 urban areas and 1 rural area in Baden-Wurttemberg: GM: 0.22 [micro]g/l (n: 343) versus 0.1 [micro]g/l (n = 71); 0.144 [micro]g/l (n = 59); and 0.165 [micro]g/l (n = 57). Among children from Sao Paulo, Brazil, HCB was detected in 19% of a sample of 251 (n = 47). (11) The mean serum concentration in the 47 children was 0.28 [micro]g/l (n = 242). When we accounted for the small group that had a detectable concentration, HCB appeared higher in our sample. Regarding the GM of DDE (0.32 [micro]g/l), no published data were available from other regions in Germany. In Michigan, where children were exposed to PCB-contaminated silos, the mean of DDT in sera was 1.5 [micro]g/l (n = 25). (12) In the study from Sao Paulo, Brazil, DDE was detected in 30% of the children. In this group, the mean DDE serum value was 0.85 [micro]g/l (n = 73). Given that only the upper 30% of the children were from South-Hessen, Germany, the mean DDE concentration was 0.75 [micro]g/l, which was comparable to the value of the children from Brazil.

The results of our study only partially supported the assumption of Dewailly et al., (4) who indicated that OCs constituted a risk for otitis media, and the assumption posited by Yu et al., (13) who indicated that OCs were risk factors for upper respiratory tract infections or for a higher susceptibility to infections. (14) PCBs and HCB showed an increased relative risk for otitis media for children in the upper half of the DDE exposure (Table 5). The findings of a Dutch study included no observation of an association between pre- or postnatal PCB exposure and rhinitis, bronchitis, tonsilitis, and otitis during the first 18 mo following birth. (15) This finding might be explained by the fact that 1 indicator of OC exposure, such as PCB, is insufficient for disentanglement of the mix of different effects of PCBs, DDE, and HCB. However, in another report, we were once again unable to find an association between PCBs and lymphocyte count (3)–which comported with the results from the study in Yu-Cheng, Taiwan (13) and with the earlier-described Dutch study. (13,15)

With respect to the etiology of atopic manifestations, one school of thought has proposed that earlier infections of the respiratory system were risk factors for asthma. (16-18) We, therefore, also included each of the 3 infectious diseases in the logistic regression models for asthma and for IgE ([greater than or equal to] 200 kU/l). The OR of having a DDE blood concentration of [greater than or equal to] 0.3 [micro]g/l vs. < 0.3 [micro]g/l increased from 3.71 to 4.19 [micro]g/l, respectively (95% CI = 1.78, 15.0; Table 5). The relative risk of DDE for IgE levels of [greater than or equal to] 200 kU/l remained unchanged.

However, the time order may also have been reversed; infections may have been reduced in favor of autoimmune responses, which are determined at birth. (19) The latter possibility was suggested by the DDE pattern of an increase of atopic manifestations and a decrease in infections (Table 6).

We considered that susceptibilities to infectious diseases and to atopic disorders were 2 opposite types of immune responses. The former was supported by Th1 and the latter by Th2. We assumed that suppression of a normal immune response against infectious diseases was related to a stronger Th2 cell response. (7,20) In addition, the production of IgE was dependent on the Th1/Th2 balance. (21) Surprisingly, DDE, HCB, and PCBs seemed to be involved in this pattern. DDE, in particular, appeared to be associated with a lower frequency of infections and with a higher prevalence of allergic reactions.

We could not find any studies–of humans or animals–about the risks of organochlorines (e.g., PCBs, HCB, DDE) for atopic manifestations. A review on research needs for endocrine disruptors, however, suggested the existence of associations between OCs and autoimmune syndromes. Results of a recent study from Slovakia (22) showed a direct association between in-utero exposure to polychlorinated dioxins and furans and cord blood IgE–a marker of atopic response. (5)

Inasmuch as we suggested that there was an indirect effect of DDE on the immune response, our findings required a better understanding of 2 issues: (1) Were endocrines involved in the regulation of immune susceptibility, and, if so, how? (2) How could DDE affect endocrine regulation? Evidence for the first question was based on observation (e.g., boys showed a higher prevalence of asthma than girls). (23,24) In addition, Xu et al. (25) showed, for Finnish adult offspring, that age at menarche of their mothers was a predictor of atopy. On the basis of animal experimental studies, investigators (20,26) suggested that sex hormones interfere with the Th1 and Th2 pattern of response. Evidence for an endocrine effect of DDE mainly stems from experimental studies. Danzo showed that DDE altered the binding to the estrogen and androgen receptors. (27) Sohoni et al. (28) reported an antiandrogenic activity of DDE. Reportedly, DDE, PCBs, and HCB have endocrine effects. (29-31) These findings supported the idea that DDE might act as an endocrine disrupter and might increase the susceptibility for an atopic response (i.e., predominance of Th2 cells).

In summary, the results of this study indicate that PCB and high-level HCB exposures in childhood are only weakly associated with otitis media, pneumonia, and pertussis. However, DDE was identified as a predictor for asthma and increased IgE blood levels. DDE may be an important factor for the increase in airway allergic diseases during the recent decades. Future research is needed if we are to evaluate the association of DDE and atopy and to disentangle the immunological mechanism involved.

Table 1.–Descriptive Characteristics of the Children’s

Cohort

Subgroup (%)

Total with OC

group (%) determination

Characteristic (N = 671) (n = 343)

Girls 46.2 43.1

Age (yr)

7 42.3 45.6

8 53.4 50.3

9-10 4.3 4.1

Breast-fed 80.8 85.1

Passive smoking in the

child’s home during the

past 12 mo

(cigarettes/day)

1-10 23.1 24.8

11-20 14.2 5.3

21-30 6.3 1.8

> 30 3.7 0.6

Has a doctor ever

diagnosed your child with:

Otitis media?

Once? 22.1 20.7

Twice? 31.0 39.1

Whooping cough? 39.9 37.6

Pneumonia? 20.2 18.5

Has your child ever had

asthma? 4.3 5.0

Did your child in the past

12 mo suffer from

asthma-like symptoms

([1] wheezing, [2] cough/

shortness of breath early

in the morning or at night,

[3] cough after exercise/

exposure to cold air or fog,

[4] attacks of dyspnea)?

One symptom? 14.9 14.0

Two and more symptoms? 10.0 10.2

Number of pertussis

vaccinations indicated on

the child’s vaccination

card

1-2 1.9 1.5

[greater than or equal to] 3 9.4 10.0

IgE serum concentration (kU/l)

Geometric mean 38.9

5-95% values 2.6-1,586.7

Note: Ig = immunoglobulin, and OC = organochlorine

compound.

Table 2.–Concentration of Organochlorine Compounds in Whole Blood

([micro]g/l) and Their Rank Correlation * (n = 343)

DDE

Geometric 95%

OC mean Median value Maximum r p

HCB 0.22 0.21 0.52 2.49 .57 < .0001

DDE 0.32 0.29 0.97 4.02

[SIGMA] PCBs

([dagger]) 0.49 0.48 1.47 4.48

[beta]-HCH 0.06 0.06 0.27 4.48

[gamma]-HCH 0.02 0.02 0.06 0.19

[SIGMA]

PCBs

([dagger]) [beta]-HCH [gamma]-HCH

OC r p r p r p

HCB .82 .0001 .55 .0001 .09 .09

DDE .62 .0001 .73 .0001 .18 .0006

[SIGMA] PCBs

([dagger]) .54 .0001 .11 .047

[beta]-HCH .14 .01

[gamma]-HCH

Notes: HCB = hexachlorobenzene, DDE = dichlorodiphenyldichloroethene,

PCBs = polychlorinated biphenyls, HCH = hexachlorocyclohexane, and

OC = organochlorine compound.

* Spearman’s.

([dagger]) Sum of PCB congeners: PCB 118, 138, 153, 170, 180, 183,

and 187.

Table 3.–Association between the 5 Different

Organochlorine Compounds (n = 343)

DDE ([micro]g/l)

OC n=81 n=92 n=83 n=87

HCB [less than 0.21- 0.30- 0.44-

([micro]g/l) or equal 0.29 0.43 4.02

to] 0.2

DDE (%)

[less than or

equal

to] 0.15

[micro]g/l,

n = 87 54.3 25.0 10.8 12.6

0.16-0.20

[micro]g/l,

n = 80 30.9 31.5 22.9 8.1

0.21-0.27

[micro]g/l,

n = 89 11.1 34.8 36.1 20.7

0.28-2.49

[micro]g/l,

n = 87 3.7 8.7 30.1 58.6

Rank correla-

tion

r .53

p .0001

Sum of PCBs ([micro]g/l) *

OC n=84 n=88 n=86 n=85

HCB [less than 0.31- 0.49- 0.77-

([micro]g/l) or equal 0.48 0.75 4.48

to] 0.3

Sum of PCBs (%)

[less than or

equal

to] 0.15

[micro]g/l,

n = 87 75.0 18.2 7.0 2.4

0.16-0.20

[micro]g/l,

n = 80 20.2 48.9 19.8 3.5

0.21-0.27

[micro]g/l,

n = 89 4.8 28.4 45.4 24.7

0.28-2.49

[micro]g/l,

n = 87 0 4.6 27.9 69.4

Rank correla-

tion

r .76

p .0001

[beta]-HCH ([micro]g/l)

OC n=58 n=110 n=86 n=89

HCB [less than 0.04- 0.06- 0.09-

([micro]g/l) or equal 0.005 0.08 4.48

to] 0.03

[beta]-HCH (%)

[less than or

equal

to] 0.15

[micro]g/l,

n = 87 72.4 18.2 12.8 15.7

0.16-0.20

[micro]g/l,

n = 80 17.2 46.4 12.8 9.0

0.21-0.27

[micro]g/l,

n = 89 6.9 31.8 44.2 13.5

0.28-2.49

[micro]g/l,

n = 87 3.5 3.6 30.2 61.8

Rank correla-

tion

r .55

p .0001

[gamma]-HCH

([micro]g/l)

OC n=95 n=135 n=113

HCB < 0.02 0.02 0.03-

([micro]g/l) 0.19

[gamma]-HCH (%)

[less than or

equal

to] 0.15

[micro]g/l,

n = 87 28.4 25.2 23.0

0.16-0.20

[micro]g/l,

n = 80 24.2 21.5 24.8

0.21-0.27

[micro]g/l,

n = 89 28.4 26.7 23.0

0.28-2.49

[micro]g/l,

n = 87 19.0 26.7 29.2

Rank correla-

tion

r .07

p .20

Notes: HCB = hexachlorobenzene,

DDE = dichlorodiphenyldichloroethene,

PCBs = polychlorinated biphenyls,

HCH = hexachlorocyclohexane, and

OC = organochlorine compound.

* Sum of PCB congeners: PCB 118, 138,

153, 170, 180, 183, and 187.

Table 4.–Prevalence of Infectious and Atopic Manifestations in

Children with Different Dichlorodiphenyldichloroethene (DDE),

Hexachlorobenzene (HCB), and [SIGMA]-Polychlorinated Biphenyls

([SIGMA]PCBs) Blood Concentrations (%) *

Whoo-

Categorical combinations of organochlorine Otitis Pneu- ping

compounds media monia cough

DDE < 0.3 [micro]g/l

(n = 173) 63.6 20.2 41.6

DDE [greater than or

equal to] 0.3

[micro]g/l

(n = 170) 51.8 16.5 33.5

DDE < 0.3 [micro]g/l HCB [less than or equal

to] 0.2 [micro]g/l

(n = 121) 65.3 20.7 41.3

HCB > 0.2 [micro]g/l

(n = 52) 73.1 19.2 42.3

DDE [greater than or HCB [less than or equal

equal to] 0.3 to] 0.2 [micro]g/l

[micro]g/l (n = 46) 37.0 13.0 21.7

HCB > 0.2 [micro]g/l

(n = 124) 57.3 17.7 37.9

DDE < 0.3 [micro]g/l [SIGMA]PCB [less than

or equal to] 0.2

[micro]g/l (n = 128) 68.0 19.5 37.5

[SIGMA]PCB > 0.2

[micro]g/l (n = 45) 66.7 22.2 53.3

DDE [greater than or [SIGMA]PCB [less than

equal to] 0.3 or equal to] 0.2

[micro]g/l [micro]g/l (n = 44) 31.8 20.5 18.2

[SIGMA]PCB > 0.2

[micro]g/l (n = 126) 58.7 15.1 38.9

IgE

[greater

than or

Categorical combinations of organochlorine equal to]

compounds Asthma 200 kU/l

DDE < 0.3 [micro]g/l

(n = 173) 3.5 12.1

DDE [greater than or

equal to] 0.3

[micro]g/l

(n = 170) 6.5 19.4

DDE < 0.3 [micro]g/l HCB [less than or equal

to] 0.2 [micro]g/l

(n = 121) 4.1 13.2

HCB > 0.2 [micro]g/l

(n = 52) 1.9 9.6

DDE [greater than or HCB [less than or equal

equal to] 0.3 to] 0.2 [micro]g/l

[micro]g/l (n = 46) 8.7 26.1

HCB > 0.2 [micro]g/l

(n = 124) 5.7 16.9

DDE < 0.3 [micro]g/l [SIGMA]PCB [less than

or equal to] 0.2

[micro]g/l (n = 128) 4.7 12.5

[SIGMA]PCB > 0.2

[micro]g/l (n = 45) 0 11.1

DDE [greater than or [SIGMA]PCB [less than

equal to] 0.3 or equal to] 0.2

[micro]g/l [micro]g/l (n = 44) 6.8 20.5

[SIGMA]PCB > 0.2

[micro]g/l (n = 126) 6.4 19.1

Note: Ig = immunoglobulin.

* Minimum values are one-half of the detection limits (detection

limits = 0.02 [micro]g for DDE, HCB, and PCB, respectively). The

maximum values of the categorized levels for DDE, HCB, and

[SIGMA]PCB are provided in Table 3.

Table 5.–Odds Ratios (ORs) and 95% Confidence Intervals (CIs) of

Dichlorodiphenyldichloroethene (DDE) for the Prevalence of Infectious

Diseases and Atopic Manifestations in Children, Controlling for Age,

Gender, and Breast-Feeding * (n = 343)

Otitis media Pneumonia

Organochlorine compound OR 95% CI OR 95% CI

DDE [greater than or

equal to] 0.3

[micro]g/l vs. < 0.3

[micro]g/l 0.5 0.31, 0.79 0.79 0.45, 1.40

Gender: female 1.23 0.78, 1.96 1.02 0.58, 1.79

Breast-fed 1.54 0.81, 2.93 1.04 0.47, 2.31

Age: [greater than or

equal to] 8 yr vs.

7 yr 0.72 0.44, 1.20 1.20 0.64, 2.25

For DDE < 0.3 [micro]g/l

([double dagger])

HCB > 0.2 [micro]g/l vs.

[less than or equal to]

0.2 [micro]g/l 1.25 0.6, 2.62 0.98 0.42, 2.56

For DDE [greater than or

equal to] 0.3

[micro]g/l ([double

dagger])

HCB > 0.2 [micro]g/l vs.

[less than or equal to]

0.2 [micro]g/l 2.38 1.08, 5.25 1.64 0.56, 4.81

For DDE < 0.3 [micro]g/l

([double dagger])

[SIGMA]PCB > 0.48

[micro]g/l vs. [less

than or equal to] 0.48

[micro]g/l ([section]) 0.85 0.4, 1.80 1.24 0.53, 2.92

For DDE [greater than or

equal to] 0.3

[micro]g/l ([double

dagger])

[SIGMA]PCB > 0.48

[micro]g/l vs. [less

than or equal to] 0.48

[micro]g/l ([section]) 3.70 1.64, 8.34 0.68 0.26, 1.76

Whooping cough

([dagger]) Asthma

Organochlorine compound OR 95% CI OR 95% CI

DDE [greater than or

equal to] 0.3

[micro]g/l vs. < 0.3

[micro]g/l 0.61 0.38, 0.99 3.71 1.10, 12.56

Gender: female 1.13 0.70, 1.84 0.17 0.04, 0.78

Breast-fed 1.84 0.88, 3.82 0.27 0.07, 0.96

Age: [greater than or

equal to] 8 yr vs.

7 yr 0.67 0.40, 1.13 1.79 0.48, 6.69

For DDE < 0.3 [micro]g/l

([double dagger])

HCB > 0.2 [micro]g/l vs.

[less than or equal to]

0.2 [micro]g/l 1.23 0.60, 2.52 1.21 0.11, 13.82

For DDE [greater than or

equal to] 0.3

[micro]g/l ([double

dagger])

HCB > 0.2 [micro]g/l vs.

[less than or equal to]

0.2 [micro]g/l 1.69 0.68, 4.20 0.53 0.13, 2.16

For DDE < 0.3 [micro]g/l

([double dagger])

[SIGMA]PCB > 0.48

[micro]g/l vs. [less

than or equal to] 0.48

[micro]g/l ([section]) 2.03 0.95, 4.30 4 cases, no

valid estimation

For DDE [greater than or

equal to] 0.3

[micro]g/l ([double

dagger])

[SIGMA]PCB > 0.48

[micro]g/l vs. [less

than or equal to] 0.48

[micro]g/l ([section]) 2.48 0.94, 6.57 0.56 0.13, 2.52

IgE [greater

than or equal to]

200 kU/l

Organochlorine compound OR 95% CI

DDE [greater than or

equal to] 0.3

[micro]g/l vs. < 0.3

[micro]g/l 2.28 1.20, 4.31

Gender: female 0.65 0.34, 1.22

Breast-fed 0.61 0.27, 1.41

Age: [greater than or

equal to] 8 yr vs.

7 yr 0.74 0.39, 1.41

For DDE < 0.3 [micro]g/l

([double dagger])

HCB > 0.2 [micro]g/l vs.

[less than or equal to]

0.2 [micro]g/l 0.91 0.30, 2.73

For DDE [greater than or

equal to] 0.3

[micro]g/l ([double

dagger])

HCB > 0.2 [micro]g/l vs.

[less than or equal to]

0.2 [micro]g/l 0.49 0.19, 1.23

For DDE < 0.3 [micro]g/l

([double dagger])

[SIGMA]PCB > 0.48

[micro]g/l vs. [less

than or equal to] 0.48

[micro]g/l ([section]) 1.08 0.35, 3.31

For DDE [greater than or

equal to] 0.3

[micro]g/l ([double

dagger])

[SIGMA]PCB > 0.48

[micro]g/l vs. [less

than or equal to] 0.48

[micro]g/l ([section]) 0.82 0.32, 2.08

Note: HCB = hexachlorobenzene, PCB = polychlorinated biphenyls, and

Ig = immunoglobulin.

* Minimum values are one-half of the detection limits (detection

limits = 0.02 [micro]g for DDE, HCB, and PCBs, respectively). The

maximum values of the categorized levels for DDE, HCB, and

[SIGMA]PCBs are provided in Table 3.

([dagger]) Including pertussis vaccination as a predictor. For the

model with DDE as a risk factor, the OR is 0.46 (95% CI: 0.28, 0.73).

* Statistically controlling for gender, breast-feeding, and age.

See the model for DDE as the only organochlorine compound

(first 2 rows).

([section]) Sum of PCB congeners: PCB 118, 138, 153, 170, 180, 183,

and 187.

Table 6.–Relationship between Blood Concentrations of

Organochlorine Compounds (OCs) and IgE Serum Levels in

Multiple Linear Regression Models (n = 319)

Group GM of IgE

OC/characteristic ([micro]g/l) n (kU/l) p

HCB ([micro]g/l) [less than or equal 87 27.5 —

to] 0.15

0.16-0.20 80 32.0 .68

0.21-0.27 89 33.9 .61

0.28-2.49 87 46.6 .26

DDE ([micro]g/l) [less than or equal 81 21.1 —

to] 0.2

0.21-0.29 92 26.5 .49

0.30-0.43 83 40.0 .08

0.44-4.02 87 62.1 .011

[SIGMA]PCBs [less than or equal 84 46.7 —

([micro]g/l, to] 0.3

7 congeners) * 0.31-0.48 88 45.1 .93

0.49-0.75 86 29.5 .28

0.77-4.48 85 22.3 .14

[beta]-HCH [less than or equal 58 32.6 —

([micro]g/l) to] 0.03

0.04-0.005 110 36.7 .75

0.06-0.08 86 32.5 .99

0.09-4.48 89 35.6 .84

[gamma]-HCH < 0.02 95 31.4 —

([micro]g/l) 0.02 135 34.6 .72

0.03-0.19 113 37.2 .54

Gender Males 182 47.3 .0042

Females 137 24.9 —

Smoking in the 0 215 25.9 —

private home 1-10 79 32.2 .40

(cigarettes/ 11-20 18 48.7 .17

day) 21-30 5 54.3 .40

> 30 2 21.6 .89

Breast-fed No 48 37.4 .58

Yes 271 31.5 —

Age (yr) 7 96 36.7 —

8 205 34.3 .78

[greater than or 18 32.1 .79

equal to] 9

Notes: HCB = hexachlorobenzene, DDE = dichlorodiphenyldichloroethene,

PCBs = polychlorinated biphenyls, HCH = hexachlorocyclohexane,

Ig = immunoglobulin, and GM = geometric mean.

* Sum of PCB congeners: PCB 118, 138, 153, 170, 180, 183,

and 187.

The authors express thanks to Dr. Rauterberg (Medical, Alimentary and Veterinary Institute for Research Middle Hessen, Division of Human Medicine, Dillenburg, Germany) for the analyses of immunoglobulin E. We are deeply grateful for the field work of the team in the NORDIG Institute in Hamburg, Germany.

This study was authorized and supported by the Ministry of Environment, Energy, Youth, Family and Health Hessen, Germany.

Submitted for publication August 21, 2000; revised; accepted for publication January 17, 2001.

Requests for reprints should be sent to Dr. Wilfried Karmaus, Department of Epidemiology, Michigan State University, 4660 S. Hagadorn Road, Suite 600, East Lansing, MI 48823.

References

(1.) Holdke B, Karmaus W, Kruse H. Body burden of polychlorinated biphenyl compounds in whole blood of 7-10-year-old children in the area of a hazardous waste incineration facility [in German]. Gesundheitswesen 1998; 60:505-12.

(2.) Osius N, Karmaus W, Kruse H, et al. Exposure to polychlorinated biphenyls and levels of thyroid hormones in children. Environ Health Perspect 1999; 107:843-49.

(3.) Karmaus W, Huang S, Osius N, et al. Chromium urine concentration and effect on lymphocyte subpopulations in children. J Environ Med 1999; 1:153-61.

(4.) Dewailly E, Ayotte P, Bruneau S, et al. Susceptibility to infections and immune status in Inuit infants exposed to organochlorines. Environ Health Perspect 2001; 108:205-11.

(5.) Reichrtova E, Ciznar P, Prachar V, et al. Cord serum immunoglobulin E related to the environmental contamination of human placentas with organochlorine compounds. Environ Health Perspect 1999; 107:895-99.

(6.) Renz H. The central role of T cells in allergic sensitization and IgE regulation. Exp Dermatol 1995; 4:173-82.

(7.) Del Prete G. The concept of type-1 and type-2 helper T cells and their cytokines in humans. Int Rev Immunol 1998; 16:427-55.

(8.) Weiland SK, von Mutius E, Hirsch T, et al. Prevalence of respiratory and atopic disorders among children in the East and West of Germany five years after unification. Eur Respir J 1999; 14: 862-70.

(9.) Statistical Analysis System, Version 8. Cary, NC: SAS Institute, 2000.

(10.) Ministerium fur Arbeit Gesundeit und Sozialordnung Baden-Wurttemberg. In: Zusammenarbeit mit dem Landesgesundheitsamt Baden-Wurttemberg und dem Statistischen Landesamt Baden-Wurttemberg (Eds). Pilotprojekt Beobachtungsgesundheitsamter. Teil 1. Stuttgart, Germany: Ministerium fur Arbeit, Gesundheit and Sozialordnung Baden-Wurttemberg; 1994.

(11.) Santos Filho E, de Souza e Silva R, Barretto HH, et al. Heavy metal and organochlorine pesticide blood levels in children of 1 to 10 years of age. Rev Saude Publica 1993; 27:59-67.

(12.) Schantz SL, Jacobson JL, Humphrey HE, et al. Determinants of polychlorinated biphenyls (PCBs) in the sera of mothers and children from Michigan farms with PCB-contaminated silos. Arch Environ Health 1994; 49:452-58.

(13.) Yu ML, Hsin JW, Hsu CC, et al. The immunologic evaluation of the Yucheng children. Chemosphere 1998; 37:1855-65.

(14.) Ross PS, De Swart RL, Reijinders PJ, et al. Contaminant-related suppression of delayed-type hypersensitivity and antibody responses in harbor seals fed herring from the Baltic Sea. Environ Health Perspect 1995; 103:162-67.

(15.) Weisglas-Kuperus N, Sas TC, Koopman-Esseboom C, et al. Immunologic effects of background prenatal and postnatal exposure to dioxins and polychlorinated biphenyls in Dutch infants. Pediatr Res 1995; 38:404-10.

(16.) Ferrari M, Tardivo S, Zanolin ME, et al. Serious childhood respiratory infections and asthma in adult life. A population-based study. ECRHS Italy. European Community Respiratory Health Survey. Ann Allergy Asthma Immunol 1999; 83:391-96.

(17.) Nystad W, Skrondal A, Magnus P. Day-care attendance, recurrent respiratory tract infections and asthma. Int J Epidemiol 1999; 28:882-87.

(18.) Martinez FD. Role of respiratory infection in onset of asthma and chronic obstructive pulmonary disease. Clin Exp Allergy 1999; 2:53-58.

(19.) Prescott SL, Macaubas C, Holt BJ, et al. Transplacental priming of the human immune system to environmental allergens: universal skewing of initial T-cell responses toward the Th2 cytokine profile. J Immunol 1998; 160:4730-37.

(20.) Whitacre CC, Reingold SC, O’Looney PA. A gender gap in autoimmunity. Science 1999; 283:1277-78.

(21.) Holt PG. Programming for responsiveness to environmental antigens that trigger allergic respiratory disease in adulthood is initiated during the perinatal period. Environ Health Perspect 1998; 3:795-800.

(22.) Kavlock RJ, Daston GP, DeRosa C, et al. Research needs for the risk assessment of health and environmental effects of endocrine disruptors: a report of the U.S. EPA-sponsored workshop. Environ Health Perspect 1996; 4:715-40.

(23.) Sears MR, Burrows B, Flannery EM, et al. Atopy in childhood. I. Gender and allergen-related risks for development of hay fever and asthma (see comments). Clin Exp Allergy 1993; 23:941-48.

(24.) Kuehr J, Frischer T, Karmaus W, et al. Clinical atopy and associated factors in primary-school pupils. Allergy 1992; 47:650-55.

(25.) Xu B, Pekkanen J, Jarvelin MR, et al. Maternal infections in pregnancy and the development of asthma among offspring. Int J Epidemiol 1999; 28:723-27.

(26.) Wilder RL. Hormones, pregnancy, and autoimmune diseases. Ann NY Acad Sci 1998; 840:45-50.

(27.) Danzo BJ. Environmental xenobiotics may disrupt normal endocrine function by interfering with the binding of physiological ligands to steroid receptors and binding proteins. Environ Health Perspect 1997; 105:294-301.

(28.) Sohoni P, Sumpter JP. Several environmental oestrogens are also anti-androgens. J Endocrinol 1998; 158:327-39.

(29.) Sonnenschein C, Soto AM. An updated review of environmental estrogen and androgen mimics and antagonists. J Steroid Biochem Mol Biol 1998; 65:143-50.

(30.) Hansen LG. Stepping backward to improve assessment of PCB congener toxicities. Environ Health Perspect 1998; 106:171-89.

(31.) Kelce WR, Stone CR, Laws SC, et al. Persistent DDT metabolite p,p’-DDE is a potent androgen receptor antagonist. Nature 1995; 375:581-85.

WILFRIED KARMAUS Department of Epidemiology Michigan State University East Lansing, Michigan

JOACHIM KUEHR University Children’s Hospital University of Freiburg Freiburg, Germany

HERMANN KRUSE Institute of Toxicology Christian-Albrecht University Kiel, Germany

COPYRIGHT 2001 Heldref Publications

COPYRIGHT 2002 Gale Group