Pertussis vaccination in infancy and asthma or allergy in later childhood: birth cohort study

Open Access
Publisher Website
Google Scholar
Abstract
Participants were the 13 971 children who survived to 1 year in the Avon longitudinal study of parents and children. The study method has been described previously,5 and details can be found on the study website (http://www.alspac.bris.ac.uk/). We obtained the vaccination status for each child from the child health surveillance database. We categorised children with regard to pertussis as fully vaccinated (completed a primary course of diphtheria, tetanus, and pertussis vaccination), partially vaccinated (completed a primary course of diphtheria and tetanus but did not receive pertussis vaccine) or non-vaccinated (no vaccinations). We excluded other combinations from analysis. We obtained three wheezing outcomes based on parental self report questionnaires (asthma at 69-81 months, wheeze with whistling on the chest at 69-81 months, and asthma diagnosed by a doctor at 91 months) and one atopy outcome based on skin prick tests at age 7 years. We defined atopy as one or more positive reactions (wheal  2 mm) to a panel of three common allergens. We selected several variables as potential confounders of the relation between exposure and outcome, which were, however, not considered to be in the causal pathway. These were, from mother's questionnaire data: maternal education, maternal smoking during pregnancy, maternal history of asthma or eczema, maternal financial difficulties, damp housing, overcrowding, child's ethnicity, number of siblings, contact with cats in the home, duration of breast feeding, and passive exposure to tobacco smoke; and, from medical records: birth weight, sex, gestational age, and maternal age at delivery. We used Pearson's χ2 (or Fisher's exact test if the predicted number of subjects in any category was less than five) for our data analysis of univariable associations between vaccination status and possible confounders and principal outcomes. We used multivariable logistic regression models to evaluate associations between immunisation status and asthma and allergy outcomes while controlling for potential confounders.