Childhood Vaccinations and Risk of Asthma

Frank Destefano, MD; David Gu, PhD; Piotr Kramarz, MD; Benedict I. Truman, MD; Michael F. Iademarco, MD; John P. Mullooly, PhD; Lisa A. Jackson, MD; Robert L. Davis, MD; Steven B. Black, MD; Henry R. Shinefield, MD; S. Michael Marcy, MD; Joel I. Ward, MD; Robert T. Chen, MD; The Vaccine Safety Datalink Research Group

Pediatr Infect Dis J. 2002;21(6) 

In This Article


Prior evidence of a possible association between vaccination and asthma was limited primarily to whole cell pertussis vaccine. The strongest association was found in a large, prospective study in New Zealand, in which none of the unvaccinated children developed asthma by age 10 years compared with nearly one-fourth of the vaccinated children.[4] The study was limited, however, by the small number of unvaccinated children (n = 23) and uncertainty about differences in medical care utilization between the two groups. A record review of 1934 patients registered from birth until at least 12 years of age in one general medical practice in the UK found a statistically significant relative risk of developing asthma of 1.44 associated with whole cell pertussis vaccine.[7] Another study of 448 English children found that 10.7% of the 243 who had received pertussis vaccine had asthma compared with 2.0% of the 203 children who had not been immunized.[5] This study was retrospective and subject to recall bias. An analysis of data from the Third National Health and Nutrition Examination Survey, involving 13 944 children, found that children who had received DTP or tetanus vaccination had a 2-fold increased risk of asthma compared with unvaccinated children.[8] This cross-sectional survey also was subject to recall bias. As part of a large randomized clinical trial of the efficacy of whole cell pertussis and acellular pertussis vaccines in Sweden, 669 children were evaluated for development of atopic disease, including asthma, from age 2 months to 2.5 years.[9] The incidence of asthma, as well as other atopic diseases, was similar in the groups of children who received whole cell pertussis vaccine, acellular pertussis vaccines and no pertussis vaccine (i.e. diphtheria-tetanus). Methodologically this was the strongest study, but the sample size was relatively small and follow-up was only to age 2.5 years. A recently reported longitudinal study from England found no association between pertussis vaccination and wheezing illnesses.[21]

A hypothesis has been developed that asthma is caused by an immunologic imbalance in the antigen-stimulated cytokine response of two classes of T helper cells (Th1 and Th2) in favor of the Th2 response.[16,22] Some studies suggest that early infections with certain respiratory pathogens may shift the Th1/Th2 balance toward Th2 cells,[23,24] whereas infections that stimulate predominantly a Th1 response (e.g. mycobacteria, measles) may decrease atopic manifestations.[15,25] This theory, however, is probably overly simplistic, and its applicability to human exposure to vaccine antigens is uncertain. Current laboratory evidence on the effect of vaccines in general and pertussis-containing vaccines in particular on Th1/Th2 balance is inconclusive. Alum, an adjuvant in vaccines against pertussis, may stimulate induction of a Th2-like response.[26,27] Pertussis toxin itself can act as an adjuvant, but it seems to stimulate a mixed Th1 and Th2 cytokine response.[28] More recent studies suggest that whole cell pertussis vaccine stimulates primarily a Th1 response.[29,30]

MMR was the only vaccine, other than pertussis, for which we had any a priori suspicion of a possible association with asthma. A study in Guinea Bissau[15] found that children who contract clinical measles may be less likely to develop atopy than children who do not contract measles (usually because they have been vaccinated). Thus measles in infancy may be protective against developing atopic conditions, including asthma, and by preventing measles disease, measles vaccine may indirectly increase the risk of asthma or atopy. These epidemiologic conclusions conflict with a recent large study in Finland, which found a positive association between measles and atopic conditions.[31] A study in a general medical practice in the UK did not find an association between measles vaccination and atopy or asthma.[7] The immune responses to measles and measles vaccine are complex, involving both Th1 and Th2 responses.[32,33,34] Our results indicate that children vaccinated with MMR are not at increased risk of developing asthma compared with unvaccinated children.

In our main analysis we found that Hib and hepatitis B vaccines were associated with 18 and 20% increases in asthma risk, respectively. The Hib results, however, were not consistent by HMO. An increased risk was evident only at NCK, and at the other HMOs the relative risks associated with Hib were <1.0. The Hib vaccines used at NCK were also used at the other HMOs; thus we do not have a ready explanation for the idiosyncratic Hib results at NCK and suspect that they are probably a random chance finding.

We are not aware of any previous studies of asthma risk after hepatitis B vaccination or of any data suggesting an association between hepatitis B disease and asthma. The hepatitis B vaccine may contain residual brewers yeast antigens from the production process, but we are not aware of any association between brewers yeast and asthma. The immunologic response to the hepatitis B surface antigen used in the vaccine appears to be predominantly Th1.[35,36] Hepatitis B vaccine, however, is the one vaccine that in the US is often administered at birth. Because it has been shown that Th2 cells dominate at birth in children who develop atopy,[37,38] we performed a subanalysis to assess whether vaccination at birth (i.e. up to 14 days of age) carried a different risk than administering the first dose of hepatitis B vaccine at later ages, but we did not find any difference (data not shown).

Potential limitations of our study included possible misclassification of asthma status, incomplete information on potential confounding factors and relatively short follow-up. Relying on computerized information, we used several criteria to identify children with asthma. A previous study using asthma prescription criteria similar to those of ours showed that they have high sensitivity and positive predictive value for asthma.[39] We performed additional analyses in which we used different asthma case definitions, and the results were not materially different from those of the main analysis. We were able to adjust for gender and date of birth and for the racial, ethnic and socioeconomic characteristics of area of residence, but we did not have information on family history of asthma and other asthma risk factors. The few increased relative risks that we found were all <1.3 and could be highly subject to possible confounding by factors on which we did not have information.

The ages of the children in our study (18 months to 6 years) may have been too young to fully evaluate asthma risk. Recent studies, however, indicate that most children with asthma are diagnosed by the age of 5 years and that symptoms usually first appear in infancy and early childhood.[40,41] In the New Zealand study differences in asthma prevalence were apparent by age 5 years.[4] That >10% of the children in our study developed asthma also suggests that the age of the children was not a major limitation.

It was critically important in our study that children for whom no vaccinations were identified in the computerized immunization tracking systems were actually not vaccinated. To have a complete vaccination history from birth is the main reason that we restricted the analysis to children who became HMO members at birth. Nonetheless it is possible that some of the children who had no immunization record in the automated immunization tracking systems actually had been vaccinated. For example some children may have had dual health plan coverage and, although enrolled in one of the VSD HMOs, received most of their medical care through another health care provider. In this case we would not have data on their immunizations or on their medical care visits, including any for asthma. Including such children in the analysis would result in a biased lower apparent risk of asthma among “unvaccinated” children.

To evaluate the magnitude of possible medical care utilization bias, we performed a subanalysis restricted to children whom we knew were using two of the VSD HMOs (GHC and NCK) for their health care because they had made at least two medical care visits during their first year of life. In this subanalysis the relative risks for almost all of the vaccines of interest decreased, including those for Hib and hepatitis B. In another subanalysis in which we tried to reduce possible health care utilization bias by restricting the analysis to children who had received at least two OPV, two DTP and one MMR, the relative risk of asthma associated with hepatitis B vaccine was less than that found in the main analysis. We conclude from these findings that the results of our main analysis are probably biased upward and tend to overestimate the relative risks associated with vaccination.

In conclusion medical care utilization bias did seem to influence the results for Hib and hepatitis B vaccines, for which we found weak associations with asthma. Despite a similar bias that would favor finding an increased risk, we found that DTP, OPV and MMR vaccines did not increase a child's risk of developing asthma.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.