You are here:
CHILDREN'S VULNERABILITY TO ENVIRONMENTAL IMMUNOTOXICANT EXPOSURE
Impact/Purpose:
This project aims at determining the immunotoxic risk in children exposed prenatally and postnatally to polychlorinated biphenyls. Experimental animal studies with Aroclor 1254 used by EPA for calculating a Reference Dose (RfD) for PCB suggest that immunotoxicity may be critical, but current exposures, especially those from breast-feeding, greatly exceed the RfD.
Description:
The total cohort was comprised of 656 Faroese children (109 of which were funded by The Arctic Environment Program and the Danish Environmental Protection Agency). The present part of the cohort had a Neurological Optimality Scale (NOS) examination at 2 weeks of age and also had valid data on mercury (cord blood and/or maternal hair) and PHAH (maternal serum and/or milk) exposure at birth.
According to the pediatric immunization schedule in the Faroe Islands (similar to Scandinavia), the children receive three vaccinations during the first year that include the tetanus and diphtheria toxoids. A booster is then given at age 5 years. In this community, the vaccinations are normally completed by all children. The present study aimed at exploring the specific antibody concentrations against the two toxoids just before the booster vaccination, with a repeat examination about one month later, as is the usual practice in vaccinology.
Of the 547 cohort children, 485 (88.7%) participated in the examination just before age 5 years. The loss to follow-up was mainly due to the family having moved abroad or the children being afraid of needles. However, from 20 of these children, serum for antibody analysis was not obtained. Some of the children or their parents had decided in advance not to participate in the blood test, while others decided at the time of the examination. The average age of the children at the time of examination was 4.96 years (s.d., 0.06). These examinations were carried out from March 2003 to January 2005. One child had to be excluded because the 5-year vaccination had already been performed prior to the examination.
With the first blood test we aimed at obtaining 30 ml of blood from the children:
- 5 ml of blood was for mercury analysis (completed)
- 5 ml of whole blood was for hematological examination (not done due to incomplete sampling)
- 2 ml serum was for POP analysis (completed)
- 1 ml serum was to track vaccination response (completed)
- 1 ml serum was for IgE (moved to age 7 years)
- 2 x 2 ml was for CRP, IgA-M, albumin, bilirubin, tryglycerides and cholesterol (moved to age 7 years)
- In addition, a hair sample for mercury analysis was taken from the occipital area (completed)
A total of 391 children (71.5% of total cohort) returned for the follow-up examination about one month after the booster vaccination. In four cases, enough serum for antibody analysis was not obtained. The average age of the children at the time of the second examination was 5.20 years (s.d., 0.13). The age range was wider, up to 5.84 years. The time from the booster vaccination averaged 35 days (s.d., 11), with a maximum of 77 days. Seven children were examined less than two weeks after the booster vaccination (and will be deleted from the statistical analysis). This variability was due to scheduling problems related to parental work obligations, incident disease, foreign travel, and similar problems. Some children were excluded from follow-up, because their 5-year vaccination had not yet been performed.
At the re-examination, only 5 ml of blood was taken in a single vial, so that, after centrifugation, 2 ml serum was available for antibody measurements (completed).
The hair and blood samples for contaminant analyses were transferred to the analytical laboratory in Denmark. The results for the pre-booster samples are:
Contaminant |
Number |
Geometric mean |
50% range |
Total range |
Blood-Hg (μg/L) |
462 |
2.67 |
1.36-5.44 |
0.03-36.5 |
Hair-Hg (μg/g) |
485 |
1.86 |
1.32-2.47 |
0.89-9.48 |
Serum-PCB (μg/g lipid) |
464 |
1.14 |
0.71-1.92 |
0.07-9.56 |
Serum-PCB teq (pg/g lipid) |
464 |
1.24 |
0.75-2.24 |
0.1-12.2 |
The total PCB concentration was calculated as twice the sum of the concentrations of congeners CB-138, CB-153, and CB-180. In addition, the dioxin-equivalent concentrations of major mono-ortho congeners CB-105, CB-118, and CB-156, was also calculated using WHO equivalence factors.
The increased level of exposure was thereby confirmed. The PCB concentrations are mainly due to the persistence of PCB transferred from the mother during pregnancy and during lactation. The mercury is due to the child’s own intake of seafood contaminated with methylmercury, which has a much shorter elimination half-life than the highly persistent PCBs.
The serum was analyzed for specific antibodies against the two toxoids, diphtheria and tetanus. The results were as follows from the two examinations:
Antibody (IU/mL) |
Number |
Geometric mean |
50% range |
Total range |
Diphtheria (1) |
463 |
0.12 |
0.05-0.40 |
0.002-6.4 |
Diphtheria (2) |
386 |
12.4 |
6.4-26 |
0.40-200 |
Tetanus (1) |
463 |
0.22 |
0.09-0.51 |
0.004-120 |
Tetanus (2) |
387 |
32.4 |
15-90 |
0.11-610 |
The antibody concentration increased by an average of 260-fold and 510-fold for diphtheria and tetanus, respectively. The highest ratios occurred in children examined about one month after the vaccination. However, due to substantial variability, the interval since the booster was not significantly associated with the post-booster antibody concentration or the ratio. Overall, the study design was successfully carried out, although some loss to follow-up happened, and the time of the second examination was not as well defined as had been intended.
The variability of post-booster antibody concentrations can probably in part be ascribed to small differences in the interval since the booster vaccination. In addition, children probably show different rates of response to the booster. This finding is somewhat surprising, because assessment one month after the vaccination is standard practice in vaccinology. However, assessment at this time would seem less suitable for identifying adverse effects of immunotoxicant exposures.
In regard to the pre-vaccination data, the diphtheria antibody concentrations decreased at increasing PCB exposure. A doubling of the serum PCB concentration was associated with a decrease in antibody concentration of about 25%. This finding is very similar to the results from the pilot study. Mercury also showed a negative association, though most likely due to confounding with PCB. Tetanus again showed a weaker effect.
The post-booster results showed a poor association with the children’s exposure biomarkers. Even when the post-vaccination results were adjusted for the pre-vaccination results, no clear effects were seen. The statistical analyses are still ongoing.
In addition to comparing the antibody outcomes with PCB exposures assessed at different times, an opportunity also existed to examine the possible association with two integrated measures of PHAH exposures. One relied on the induction of estrogen-like properties, especially those mediated through interaction with the nuclear estrogen receptor. In addition, 2,3,7,8-tetrachlorinated dibenzo-p-dioxin (TCDD) and related substances interact with the intranuclear arylhydrocarbon receptor (AhR). Because a variety of substances may act as agonists or antagonists, assessment of complex exposures is complex, when effects are mediated via receptor interaction.
In this project, we used serum to assess the receptor activity in vitro, where activation was expressed by a reporter gene in a cell line. We used the E-screen and the lipid-adjusted aryl hydrocarbon receptor (AhR) activity as the two approaches to measure xenoestrogenic responses resulting from the combined exposure to the endocrine disrupting chemicals, and to assess the potential correlates of the xenoestrogenic activity, very little of which is known. The E-screen was conducted by stripping the serum for endogenous hormones and adding the extract to the MCF-7 breast cancer cell line, where induced growth was assessed in regard to the response induced by estradiol (E2).
The Ah receptor assay, serum or hexane-extracted serum lipophilics were added to cell line with a GFP reporter gene ligated to multiple copies of activated AhR-binding sites in the CYP1A1 promoter, and the activity was calibrated to that of TCDD. In both cases, we applied serum from the 34th week of pregnancy of the birth cohort mothers.
Increased E2 and AhR activities were found in some of the serum samples. Correlations were then assessed from lipid-based concentrations of individual organohalogens determined in the serum. The main finding was that β-HCH, p,p’DDE, and CB-118 were significantly negatively correlated with the E2 activity (r=-0.22, p=0.004; r=-0.15, p=0.05; r=-0.16, p=0.03 respectively). β-HCH and dieldrin were negatively correlated with AhR (r=-0.26, p=0.002; r=-0.23, p=0.006).
However, we failed to identify any single organohalogen substance responsible for the induced E2 and Ah activities, several negative associations were found. These findings are in accordance with the notion that the receptor activation is a result of complex interactions that are difficult to disentangle from measurements of individual agonists or antagonists. However, the results suggest that the E2 and Ah activity measure may be useful as integrated exposure indicators.
When used at statistical predictors for antibody concentrations at age 5 years, no clear pattern was revealed. However, these parameters will be included as exposure biomarkers in the structural equation models currently being developed.
All exposure biomarkers are associated with imprecision and risk of misclassification. The routine practice of comparing individual PCB congener concentrations with outcomes runs into the additional problem of multiple comparisons. We chose to avoid this problem by calculating to PCB exposure indices from the congeners most prevalent and therefore also most likely the ones determined with the greatest precision.
In further analyses of the cohort data, structural equation models will be used. Here the exposure variables will be assumed to reflect a latent, true exposure, which is not observed but can be modeled from the data. In these models, the observed PCB concentrations were considered manifestations of a latent true exposure, which, in turn, was assumed to affect each of the two antibody concentration outcomes. The latent exposure variable was assumed to depend on maternal blubber intake (dietary questionnaire response), and covariates were allowed to correlate with the latent exposure and the outcome. Unlike standard regression techniques, this approach takes into account measurement error in the exposure variables. In addition, a more powerful analysis is obtained, because information from different exposure markers is pooled.
We have used this approach in our pilot study, where we showed that it is highly feasible and powerful. The models will include sex, age, birth weight, and maternal smoking during pregnancy. Residual plots were used to assess the model fit, and regression results were also compared with the estimated dose-response functions from generalized additive models. An effort will be made to separate the relative effects of prenatal exposures (from maternal PCB concentrations) and postnatal exposures, as determined from the child’s serum concentrations at age 18 months (N ~ 100, pilot study) and 5 years.
We calculated the benchmark dose as the PCB exposure that increased the risk of an abnormal response from the anticipated 5%–10%. To take the statistical uncertainty into account, the lower 95% confidence limit of the benchmark dose was then calculated, in accordance with previously established procedures.
Technical aspects
This study benefited from the advantageous epidemiological setting in the Faroe Islands. This population is exposed to elevated concentrations of PCBs from their traditional diet, which includes blubber from the pilot whale, a small toothed whale in which persistent contaminants are biomagnified. However, some Faroese eat very little traditional food, and a few eat a lot, so PCB concentrations range over two orders of magnitude, and the average exposure exceeds the US average by about 10-fold.
An additional advantage is that prospective cohort studies have been carried out in this population for over 20 years, and that the population is served by a single-provider government-owned health care system, which facilitates follow-up. Participation rates are generally close to 90%, although with lower rates for blood sampling from children.
The study protocol called for the first examination to occur just before the booster vaccination at 5 years of age. Only in one child had this vaccination already been done by the time of the first examination. For some children, the booster was delayed, whether due to incident disease, travels, or other reasons. However, the main problem in regard to the second clinical examination was that the children had already participated once, had recently been vaccinated, and therefore were more likely to refuse an additional blood test. Still, a participation rate above 70% is quite remarkable under these circumstances.
Childhood immunizations represent a standardized challenge to the children’s immune system, a challenge that is not otherwise available or appropriate. This project focused on specific antibody reactions to protein toxoids, because they depend on the functions of several types of immune cells. The antibody assay has been standardized and is a routine analysis at national laboratories, such as the State Serum Institute in Copenhagen, Denmark.
Despite the fact that this population-based cohort of children was in good health, the antibody concentrations varied substantially. Known or suspected risk factors contributed very little to explaining this variability. In this connection, the significant effect of PCB exposure would therefore seem remarkable.
The Faroese population is exposed to PCBs at a much higher level than the US population, but the results suggest that possible adverse influences on the immune function may well occur also at lower exposure ranges prevalent worldwide. Further exploration of this issue will depend on the calculation of benchmark dose results.
Exposure assessment is always associated with a certain degree of imprecision. The use of biological samples for analysis of exposure biomarkers represents a substantial improvement, but these parameters are known to depend on a variety of factors, such as sampling time and the temporal characteristics of the exposure. These biomarkers may not necessarily reflect the exposure at target organ level at the time of the greatest vulnerability. Such imprecision will generally be non-directional and will cause a bias toward the null. In this protocol, the greatest possible effort was therefore made to control and minimize any error sources.
In regard to the PCB analysis, recoveries were >96% for the congeners measured. Within-day repeatability and between-day reproducibility showed a coefficient of variation < 8%. The quality control and accuracy is ensured by biannual participation in the external quality program (EQUAS) organized under the auspices of the German Society of Occupational Medicine and Environmental Medicine (Prof. Dr. H.Drexler, University Erlangen-Nuremberg, Erlangen, Germany). Likewise, mercury analyses are secured by participation in the Canadian external quality assurance program.
The antibody concentration assays were carried out using WHO-approved protocols, and the results were expressed in international units. The wide variability of the results is not affected by the slight analytical imprecision.
The two cell-based assays were comparatively labor-intensive. Given that neither of them showed a clear association with specific exposure parameters or outcomes, their use in future studies in this field needs to be considered with some skepticism.
Overall, this study provides supplementary epidemiological evidence for an association between exposure to environmental pollutants and a reduction of antibody production after routine childhood immunizations.
Implications
Among chemicals persistent in the environment, Polyhalogenated aromatic hydrocarbons (PHAHs), and polychlorinated biphenyls (PCBs) are listed as Priority Persistent Bioaccumulative Toxic Chemicals by the U.S. Environmental Protection Agency, and PCBs have been said to constitute the leading chemical risk from fish consumption. They are included in the Stockholm Convention. While average PCB exposure levels have decreased over the last 25 years or so, the number of fish advisories due to PCB contamination has steadily increased as knowledge on the extent of PCB pollution has expanded.
Immunotoxicity as an outcome of pollutant exposures is of considerable relevance to human health. Studies of children born preterm have shown that small differences in immune competence can result in changes in antibody production. Even if small differences may be of limited importance with regard to the protective effects of the vaccines as such, the functional deficits represent a quantitative measure of immune system functions that may relate to possible increases in the risk of infections in general. In addition, such alterations in immune function, when incurred during development of the immune system, are likely to be permanent and could therefore impact on an individual's future health and immune-related disease risk.
A further concern is that individual immunity must be reached in a certain proportion of the population to achieve 'herd immunity'. If this proportion falls below what is necessary for a particular disease, then the population as a whole loses its protected state, and an infectious epidemic may result.
This study was designed to assess whether prenatal and postnatal exposure to PCBs impacts on antibody response to childhood immunizations. The major advantages of this study were: detailed follow-up of an established and well characterized birth cohort of sufficient size examined at appropriate ages within the time period of a 3-year grant; occurrence of developmental PHAH exposure with a mixture of PCB congeners similar to the one prevalent in North America, an exposure range that spans a 100-fold interval, and an average in excess of the US level; application of sensitive and appropriate methods of documented feasibility with regard to assessment of exposures and outcomes; and a high statistical power due to the above characteristics that will allow advanced statistical analysis including structural equations and benchmark analyses.
This epidemiological study of healthy children provides evidence of pollutant-induced reduction of antibody production following routine childhood immunizations. Although subtle, such immune function impairment could become clinically important when the immune function is challenged by other factors such as preterm birth, chronic infection, or other disease. In addition, even slight impairments could be important at a population level, e.g., with regard to an increased prevalence of respiratory tract infections.
URLs/Downloads:
2004 Progress ReportURL
2003 Progress Report
2005 Progress Report
Final Progress Report