Early exposure to antibiotics and infections and the incidence of atopic eczema: A population-based cohort study


Jochen Schmitt, Department of Dermatology, University Hospital Carl Gustav Carus, Technical University Dresden, Fetscherstr. 74, D-01307 Dresden, Germany
Tel.: 0049 351 4582421
Fax: 0049 351 4585326
E-mail: jochen.schmitt@uniklinikum-dresden.de


Schmitt J, Schmitt NM, Kirch W, Meurer M. Early exposure to antibiotics and infections and the incidence of atopic eczema: A population-based cohort study.
Pediatr Allergy Immunol 2010: 21: 292–300.
© 2009 John Wiley & Sons A/S

It has been suggested that infants exposed to antibiotics are at increased risk for atopic eczema (AE), whereas the early exposure to infections might be protective. This study describes the complex relationship between early exposure to infections, anti-infectious treatment with antibiotics, and incident AE. Using a German population-based administrative health-care and prescription database, we established a cohort of 370 children not diagnosed as having AE during their first year of life. For each individual child we identified all infections and prescriptions of antibiotics within the first year as well as incident AE within the second year of life. Crude analyses suggested that early infections and exposure to antibiotics are risk factors for AE. However, stratified analyses indicated that early infections were only associated with a higher rate of AE when treated with broad-spectrum antibiotics such as cephalosporines or macrolides. The risk ratio (RR) of children with early respiratory tract infections not treated with antibiotics was 0.69 [95% confidence interval (95% CI) 0.39 to 1.24], whereas respiratory tract infections treated with macrolides (RR: 2.15, 95% CI: 1.18–3.91) or cephalosporines (RR: 1.93, 95% CI: 1.07–3.49) significantly increased the risk for AE. The results for other common childhood infections tended to be similar. Antibiotic treatment appears to modify the association between early infections and subsequent AE. We found no evidence that infections per se significantly alter the likelihood for subsequent AE.