I-Jen Wang,1,2,3 Li-Min Huang,4 Yueliang Leon Guo,5 Wu-Shiun Hsieh,4 Tien-Jen Lin,6 Pau-Chung Chen7
1Department of Pediatrics, Taipei Hospital, Department of Health, Taipei, Taiwan;
2College of Public Health, China Medical University, Taichung, Taiwan;
3 College of Medicine, Fu Jen Catholic University, Taipei, Taiwan;
4 Department of Pediatrics, National Taiwan University Hospital, Taipei, Taiwan;
5Department of Environmental and Occupational Medicine, National Taiwan University Hospital, Taipei, Taiwan;
6Taipei Medical University, Wan Fang Hospital, Taipei, Taiwan;
7Institute of Occupational Medicine and Industrial Hygiene, National Taiwan University College of Public Health, Taipei, Taiwan
Running title: Hib combination vaccines and atopic disorders Word count: 2589
Correspondence to: Professor Pau-Chung Chen
Institute of Occupational Medicine and Industrial Hygiene, National Taiwan University College of Public Health
Room 733, #17 Syujhou Road, Taipei 10055, Taiwan Phone: +886-2-3322 8088
Fax: +886-2-2358 2402 Email:[email protected]
*Manuscript
1
Background/Purpose: Epidemiologic evidence for an association between
vaccinations and atopy development is inconsistent. We evaluated the influence of Haemophilus influenzae type b (Hib) combination vaccines in 6 month old infants on
the prevalence of atopic disorders in 18 month old children.
Methods: We used multistage, stratified systematic sampling to recruit 24,200
mother-newborn pairs from the Taiwan national birth registration in 2005. Vaccination status was ascertained through official vaccine cards, while risk factors for atopic disorders were gathered by questionnaires at 6 months of age. Information about development of atopic dermatitis (AD) and recurrent wheezing was collected at 18 months of age. The relationship between atopic disorders and Hib combination vaccines, diphtheria-pertussis-tetanus-Hib and oral poliomyelitis vaccines
(DPT-Hib&OPV) and DPT-Hib-inactivated poliomyelitis vaccines (DPT-Hib-IPV), were estimated by multiple logistic regression.
Results: A total of 19,968 children completed the follow-up and participated in the study. AD was noted in 1584 (7.9%) infants while recurrent wheezing was found in 1,220 (6.1%). The adjusted ORs (95% CI) for the development of AD in the
DPT-Hib&OPV and DPT-Hib-IPV vaccination groups were given as 1.38 (1.15-1.65) and 1.49 (1.29-1.72), compared to those without Hib vaccination (DTP&OPV
vaccination). However, the association between DPT-Hib&OPV and DPT-Hib-IPV vaccinations and recurrent wheezing failed to reach statistical significance.
Conclusion: There is a potential risk for AD after receiving Hib combination vaccines.
Whether the trivial adverse impact to increased risk of atopic disorders outweighs the Hib vaccination importance in public health for infectious diseases spreading warrants further investigation.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
Key words: atopic dermatitis, recurrent wheezing, Haemophilus influenzae type b, vaccines
1 decades.1 The hygiene hypothesis has emerged, linking reduced microbial exposure and infections early in life with the development of allergic diseases.2 Early childhood immunizations have been viewed as promoters of atopic disease development by stimulating a Th2-type immune response or decreasing microbial pressure, which shifts the balance between Th1 and Th2 immunity.3 Because serious infectious diseases have been almost eradicated in some populations, whether the possible adverse effects of vaccinations may take precedence over their preventive action should be paid more attention. Since vaccination is universally recommended for infants, any association between childhood vaccinations and the occurrence of a common serious condition such as allergic diseases could be of considerable public health importance.
The possible effects of Bacillus Calmette-Gue´rin (BCG), oral poliomyelitis (OPV), measles-mumps-rubella vaccination (MMR), and diphtheria-tetanus-pertussis (DTP) vaccinations on atopic disease have been studied.4-7 Neonatal BCG vaccination was reported to be associated with a significantly lower prevalence of wheeze and asthma.4 A study in U.K. indicated a slight increase in AD risk at 12 years of age in those who had received the DTP vaccine.6 Furthermore, a Danish population-based cross-sectional study demonstrated an almost two fold increase in AD risk in children after receiving MMR vaccine.7 However, no conclusive data on the relationship of Haemophilus influenzae type b (Hib) vaccines and Hib combination vaccines with
atopic disorders are yet available because they were introduced recently to most vaccination programs.8-10 The studies suffered from some drawbacks such as information bias, selection bias, medical care utilization bias, and reverse causation
1
because of cross sectional design. One study even showed a mixed Th1–Th2 response after the Hib vaccination of 19 adults.11 Biological plausibility for a causal
relationship is not apparent. Therefore, more evidence concerning Hib vaccination as a potential risk factor for atopic disorders is necessary. In this study, we investigated the risk of atopic dermatitis and recurrent wheezing after receiving Hib combination vaccines.
1
Childhood Vaccinations in Taiwan
The Hib vaccine was first introduced into Taiwan in 1997. Although the
government authorities encourage Hib vaccinations, this vaccine has remained “at one's own expense” until now. During the time of this study, infants in their first 6 months might receive Hib vaccination either as a combined DTP-Hib vaccination and oral poliomyelitis vaccination (DPT-Hib&OPV, so called four combination vaccine) given simultaneously or a combined DPT-Hib-inactivated poliomyelitis vaccination (DPT-Hib-IPV, so called five combination vaccine) as a single injection at 2, 4, and 6 months of age. Other vaccinations that are injected before 6 months of age are
DTP&OPV, BCG, and hepatitis B vaccines (HBV), and MMR mass vaccinations in National Immunization Program. In Taiwan, each vaccination is recorded in the
child‟s health card, which has to be examined before primary school entry. Therefore, parents are required to maintain accurate health cards of their children. After
vaccination, nurses update the child‟s health card kept by parents and register this information in the government computerized immunization system. Since the computerized immunization tracking system depends on the reporting of each
vaccination by the infant welfare center, we could not rule out that vaccinations were incompletely reported. Therefore, we ascertained vaccination status in the first 6 months of life by reviewing the child‟s health card during a home interview by trained interviewers.
Study Population and Sampling Strategy
The Taiwan Birth Cohort Study is a prospective longitudinal cohort study. In the study, we used a multistage stratified systematic sampling design to obtain
representative samples from the Taiwan national birth registration data in 2005. We
1
categorized a total of 369 towns in Taiwan into 12 strata according to the
administrative division (four strata) and the total fertility rate (three strata). Using the principle of probability proportional to size, we randomly sampled 90 out of 369 towns in Taiwan. A total of 24,200 pairs of parents and newborns from these 90 towns were recruited. All study participants signed informed consent as approved by the Institutional Review Board of the National Taiwan University Hospital.
Data Collection and Case Definition
We conducted a home interview with the postpartum parents at 6 and 18 months after the delivery using a structured questionnaire as previously described.12,13 Cases of AD were defined as physician-diagnosed AD through the questions „Has your child ever had AD diagnosed by a doctor?‟, and „Has your child ever had recurrent itchy rash for at least 4 consecutive half-months periods over elbows, knees, face, wrists, or generalized (4 or more localizations)?‟12-14 Children whose mothers reported that the child ever had wheezing in the chest more than 3 episodes were classified as having recurrent wheezing.15 Questions about potential confounders and prenatal and postnatal risk factors were asked at 6 months of age, including maternal age and education, family income, maternal history of atopy, maternal smoking during
pregnancy, diet and supplements during pregnancy, duration of breast feeding, age of introduction of solid foods, number of siblings, day care, pet raising, carpets at home, environmental tobacco smoke (ETS), fungi on the house wall, and place of residence.
From the records of the collaborating hospitals, we collected neonate health data at birth, such as gender, head circumference, birth body weight, height, weeks of gestation, parity, and type of delivery.
Statistical Analysis
Vaccination status were divided into 3 categories, with DTP&OPV vaccination
1
standing for no Hib vaccination and DPT-Hib&OPV or DPT-Hib-IPV vaccination standing for receiving Hib vaccination, with the DTP&OPV vaccination used as the reference category. For the association between vaccination status with atopic
disorders, odds ratios (ORs) with 95% confidence intervals (95% CIs) were calculated by logistic regression models. Analyses were performed in both univariate and
multivariate models. Potential confounders, including infant gender, preterm birth, birth weight, maternal age and history of atopy, maternal education and nationality, maternal smoking during pregnancy, family income, duration of breast feeding, day care, number of older siblings, pet raising, fungi at house wall, carpets at home, postnatal ETS exposure, MMR and Japanese B encephalitis (JBE) vaccinations, and place of residence were taken into consideration. Variables were included in the model if they changed the univariate point estimate by at least 10%. All hypothesis testing was two-sided at the significance level of 0.05 and was performed using SAS Software Version 8.2.
1
We excluded those with loss follow-up and incomplete questionnaires, multiple gestation (twins, triplets, etc), inability to answer questions, plan to move out of the area before delivery, children with immunodeficiency, structural airway abnormalities, congenital defects, and history of severe vaccination allergy. At 18 months of age, 20,172 (83.4%) children completed the follow-up. Since almost all children received DTP, OPV, BCG, or HBV mass vaccinations in National Immunization Program, we excluded those without these vaccinations. We also restricted our analyses to infants at risk for a first episode of AD or recurrent wheezing after 6 months of age to
exclude potential reverse causation between vaccinations and atopic disorders. Finally, 19,968 (99.0%) children were included in this study (Figure 1). There were no
significant difference of the characteristics between the 204 non-participants and the 19,968 participants, including maternal age, maternal history of atopy, birth weight, gestational age, gender, and parity.
The demographic characteristics of children, mother, and environmental factors were displayed in Table 1 and Table 2. There was no significant difference among the three vaccination groups with the exception of maternal education, maternal
nationality, family income, and place of residence. Children with higher maternal education, mothers of Taiwan nationality, higher family income, and living in urban areas received more Hib combination vaccines.
During the study period, a diagnosis of AD was made in 1584 (7.9%) out of 19,968 children while recurrent wheezing was found in 1220 (6.1%) children. By the age of 6 months, 5997 (30.0%) of the infants received DPT-Hib&OPV vaccination while 8254 (41.3%) of the infants received DPT-Hib-IPV vaccination. Table 3 showed odds ratios (ORs) with 95% confidence intervals (95% CI) for the relationships between
1
vaccination status and atopic disorders. In the multivariate analysis, the adjusted ORs (95% CI) for the development of AD in the DPT-Hib&OPV and DPT-Hib-IPV vaccination groups were given as 1.38 (1.15-1.65) and 1.49 (1.29-1.72), compared to those without Hib vaccination (DTP&OPV vaccination). However, the association between DPT-Hib&OPV and DPT-Hib-IPV vaccinations and recurrent wheezing failed to reach statistical significance.
1
Compared with DTP and OPV vaccines, few studies have evaluated whether Hib vaccination has any effect on atopic disorders and the results have been inconsistent.
In this cohort study, we found a weak association between Hib combiantion vaccines and AD, compared to those without Hib vaccination. A survey of 1,943
schoolchildren aged 5-14 years in Eastern Germany also found a positive association between Hib vaccination and asthma (OR, 1.86; 95% CI, 1.05-3.32).8 On the contrary, Kummeling et al. discovered that the risk of eczema or recurrent wheeze at 1 year of age did not differ between infants with different vaccination status at the age of 6 months.10 However, infant eczema and recurrent wheezing at such young age may be fairly unspecific for atopy. Bernsen et al estimated the risk of atopic disorders at ages 8–12 years to be 1.09 (95% CI: 0.75-1.58) following Hib vaccination in the first year of life in Netherlands.9 He claimed that the result might be underestimated due to selection bias and nondifferential misclassification in this cross-sectional study. Many factors could interfere with the effect of Hib vaccination and atopy such as age and ethnicity of study population, validation of vaccination status, the vaccination schedule, and the number of inoculations.3 Furthermore, various brands of vaccines that differed in the number of antigens, protein carriers and the type of adjuvants might also account for the discrepant findings of these studies.
There are theoretical reasons to suspect a correlation of atopic disorders with Hib vaccination. One possible mechanism is that vaccination may shift the immunologic balance toward a Th2-type immune response.3,16 An IgE response to vaccine antigens is frequently detected in the sera of children after primary vaccination.17 After booster vaccination, more than 90% of children have detectable IgE against the vaccine antigens.18 Moreover, the IgE response to vaccine antigens seems to be more
1
prominent among atopic individuals.17,19 Another possibility is that vaccines or their adjuvants may have direct IgE-potentiating effects.20 Hib vaccine comes with adjuvants such as aluminum salts and oil-based emulsions to enhance antibody production. The elevated humoral response induced by aluminum salts is associated with a predominant Th2-type immune response.20 Trace amounts of streptomycin, neomycin, and polymyxin B in the vaccine may also trigger allergic reactions.20 On the other side, modern subunit vaccines may not activate dendritic cells efficiently and may impair regulation of the adaptive immune response since they often lack
microbial antigens such as lipopolysaccharide, heat shock proteins and the CpG motif of bacterial DNA.21,22 Hib vaccine is an inactivated vaccine and is unlikely to mimic a natural infection-mediated immune response that protects against the development of allergic diseases. Furthermore, infants require multiple Hib immunizations because of immature T cell-dependent antibody responses. CD4 T cell responses are usually slower to develop, less readily sustained, and more easily biased towards a Th2-type response.23 In fetal life, the immune system is skewed toward TH2-type immunity, but after birth, it becomes progressively skewed toward TH1-type responses, which are crucial for a host defense against infections.24 When infants encounter with
environmental allergens or receive many inactivated vaccines early in life, the normal progression from TH2- to TH1-type immunity after birth might be disturbed and favors detoured to the Th2-skewed pathway.24
Interestingly, we found infants with Hib vaccination was associated with AD while compared to those without Hib vaccination. However, Hib vaccination failed to reach statistically significant association with recurrent infant wheezing. Different
pathogenesis between AD and wheezing might account for this finding. Another cohort study in Taiwan, with similar prevalence of wheezing, found that recurrent
1
infant wheezing during the first 18 months of age may not be related with allergy sensitization.25 Alternatively, it was associated with the Clara cell protein 10 (CC10) G+38A polymorphism and lower CC10 levels.25 Further follow up of this cohort is warranted to clarify this problem.
Our study population was a homogeneous group and all adhered to the same religious group. Therefore, the main reason for not receiving Hib combination vaccines in our study was that the vaccination had not yet been introduced into the National Vaccination Program and it was not free. This implied that probably “family income” was an important confounder. As shown in Table 1, children with higher family income, mothers of Taiwan nationality, higher maternal education, and living in urban areas received more Hib combination vaccines. These variables were taken into consideration in the statistical analyses and were included in the final model if they changed the univariate point estimate by at least 10%. However, the statistical significance of the association between Hib vaccine and AD was retained even after adjusting for the potential confounders. In addition, medical care utilization bias did not influence our results because more than 99% of participants were covered by national health insurance in Taiwan, suggesting that bias from differential access to medical care was not possible.
Motives to abstain from vaccinations, such as infant‟s illness or fear of side effects such as allergies, might lead to the reverse causation of vaccination effects. Children with severe allergy to vaccine components or immunodeficiency failed to receive some vaccinations, however, they were excluded from our study subjects.
Furthermore, “infant‟s illness” was often a reason for delayed vaccination but not for refusal to vaccination. Children were usually brought to immunizations by their mothers. Therefore, “fear of side effects such as allergies” would only be the case if
1
fear of side effects was related to an increased risk of atopic disorders by a maternal history of atopic diseases. Remes et al. reported that allergic manifestations in parents may affect the vaccination rate of children.26 However, maternal history of atopic diseases. It had been included as a potential confounder in our statistical analyses.
Furthermore, if measurement error did occur, it tended to be toward the null and the effect of vaccination was likely to be underestimated.
One of the potential limitations in our study is the relatively short follow-up. The ages of the children in our study might be too young to fully evaluate the risk of atopic diseases. However, approximately 60% of AD cases arose in the first year of life.27 That 7.9% of the children in our study developed AD also suggested that the age of the children was not a major limitation. Furthermore, our study was limited by the use of questionnaire for ascertaining AD, which might not be accurate as the standard diagnostic criteria by a dermatologist.28 Nevertheless, maternal report of doctor-diagnosed visible eczema with typical morphology and atopic distribution has been previously validated versus clinical examination in the studies performed in the United States and Denmark and has been used by many epidemiological studies.12-14,
29, 30
The strengths of our study included the large sample size and the population based prospective cohort design, with relatively short intervals between subsequent
follow-up questionnaires, which decreased the possibility of recall bias. Because of the large sample size, we were able to control for numerous potential confounders by the statistical analysis. Therefore, we are fairly confident about our results. The multistage stratified systematic sampling including urban, suburban, and rural population could enhance generality. In addition, we used the official child‟s health card to confirm the vaccination status of each child, which increased the validity of
1
the results and decreased misclassification. We compared a subset of the health cards to the information contained in medical charts and found that the data was totally in agreement with medical charts. Therefore, we were confident about the validity of the vaccination status. Since the follow-up rate was high, selection bias could not be a major problem in our study. Moreover, the prevalence of most variables in those lost in the follow-up and in those completed the followed-up was not significantly different from each other.
In conclusion, we found there is a minimal risk for AD after receiving Hib combination vaccines. Just as Guillan-Barré syndrome and the influenza vaccine, whether the trivial adverse impact to increased risk of atopic disorders outweighs the Hib vaccination importance in public health for infectious diseases spreading warrants further investigation.
1 Bureau of Health Promotion, Department of Health (DOH94-HP-1802, DOH95-HP-1802 and DOH96-HP-1702), and the National Science Council (NSC 97-2314-B-192 -001 -MY2) of Taiwan. All the authors have declared that they have no conflict of interest. We appreciate the support and statistical assistance from Professor Tung-Liang Chiang, Institute of Health Policy and Management, National Taiwan University College of Public Health, Taipei, Taiwan; Professor Shio-Jean Lin, Department of Pediatrics, National Cheng Kung University Hospital and College of
1 Bureau of Health Promotion, Department of Health (DOH94-HP-1802, DOH95-HP-1802 and DOH96-HP-1702), and the National Science Council (NSC 97-2314-B-192 -001 -MY2) of Taiwan. All the authors have declared that they have no conflict of interest. We appreciate the support and statistical assistance from Professor Tung-Liang Chiang, Institute of Health Policy and Management, National Taiwan University College of Public Health, Taipei, Taiwan; Professor Shio-Jean Lin, Department of Pediatrics, National Cheng Kung University Hospital and College of