Protective factors in childhood allergy related to diet and lifestyle

The prevalence of allergic disease among children has increased markedly during several decades. The reasons behind the increase remain unclear, but lifestyle and environmental factors associated with the Western way of living seem to be of importance. The Western lifestyle includes changes in diet, reduced infant infections, increased use of antibiotics, antipyretics and vaccinations, all of which have been investigated in relation to allergic disease, but the role of these factors for development of allergic disease has not been clarified. The aim of this thesis was to investigate if certain dietary and lifestyle factors could have a protective effect on the development of allergic disease in childhood. It is based on two epidemiological materials, the multinational PARSIFAL study (including Steiner school children, farm children and two reference groups in five European countries), and the Swedish BAMSE birth cohort study. PARSIFAL had a crosssectional design with nearly 15000 children aged 5-13 years. BAMSE included more than 4000 children, who were followed from birth until 8 years of age. Focusing on the 4600 Steiner school children and their 2000 referents in the PARSIFAL study, we could in a European context confirm the lower prevalence of allergic disease among Steiner school children observed in a previous Swedish study. Use of antibiotics and antipyretics during the first year of life was associated with increased risks of asthma and eczema, while an increased risk of rhinoconjunctivitis was related to antibiotic use. However, we can not exclude that other factors in the anthroposophic lifestyle contributed to these results. We also studied the association between measles vaccination and measles infection and allergic disease among all children in the PARSIFAL study. To reduce disease-related modification of exposure we excluded children with early symptoms of allergic disease in some analyses. After this exclusion, no association was observed between measles vaccination and allergic disease, while measles infection was inversely related to allergic disease (OR 0.65, 95% CI 0.45-0.95). In the BAMSE cohort, 2400 children provided information on diet and allergic disease at 8 years of age. A high intake of fruits was inversely associated with rhinitis, but after considering disease-related modification of consumption, that is, excluding all children who reported food related allergic symptoms, most associations became nonsignificant. Furthermore, when we investigated the role of multivitamin supplementation no association was observed between current use and risk of allergic disease, while an inverse association was indicated between early life multivitamin supplementation and sensitization to food allergens (OR 0.61, 95% CI 0.39-0.97). It may be concluded that disease-related modification of exposure (reverse causation) is an important aspect to consider in studies on behavioral factors and allergic disease. This is especially important in cross-sectional studies, but the fact that the natural history of allergic disease might cause the same child to experience different symptoms at different ages, suggests that disease-related modification of exposure may be a possible confounding factor also in prospective studies