Atopic Dermatitis and Eczema Association Between Early Life Antibiotic Exposure and Development of Early Childhood Atopic Dermatitis

Atopic dermatitis (AD) is a chronic inflammatory skin disease that begins usually during infancy and affects about 11% of children in the United States. This study sought to examine the association between antibiotic exposure and the development of AD.

Infants at the University of Florida from June 1, 2011 to April 30, 2017 who had 2 or more well-child visits after birth, with at least 1 visit at 300 days of life or later, were included.

This study was a retrospective, observational study that analyzed 4106 infants at the University of Florida from June 1, 2011 to April 30, 2017. The electronic health record was reviewed retrospectively. Infant charts were reviewed to assess the development of AD via diagnosis codes. Additionally, medications coded as antibiotics were reviewed by the authors. Antibiotic exposures were included if they were the first exposure, if they occurred before the diagnosis of AD, and only if the patient received the medication. Prenatal antibiotic use was analyzed, but nonantibacterial medications and topical antibiotics were excluded. Regarding antibiotic exposure after birth, the first month of life was examined by weekly intervals, and then monthly intervals. Identification of AD diagnosis was based on the International Classification of Diseases, Editions 9 and 10 codes for AD. Statistical analysis was performed, including summary statistics, independent samples t-tests or χ-square tests, and logistic regression.

Study demographic groupings included gender, race, delivery mode, NICU stay, gestational age, and birth weight. Infants included in the study demonstrated diversity regarding gender (48.8% female, 51.2% male), race (39.2% black, 38.5% non-Hispanic white), delivery mode (65.0% vaginal, 35.1% Caesarean section), and gestational age (46.0% full term, 28.3% early term). Half (50.4%) of the children received antibiotics at any time during the study period and 29.7% received antibiotics in the first year of life. AD was diagnosed in 26.5% of the cohort (95% confidence interval = 25.2%–27.9%), and the mean age of diagnosis of AD was 9.45 (9.53) months.

In the first year of life, 28% of children who did not receive antibiotics and 22.3% of children who did receive antibiotics developed AD. Regarding antibiotic exposure in the prenatal period, there was no difference in rates of AD in children who were exposed to antibiotics in utero versus children who were not exposed. Of note, the study showed that children who were exposed to antibiotics at weeks of life 1, 2, 3, or 4 developed AD less than children who were not exposed. However, the difference was noted to be statistically significant for infants at specifically week 2 of life. After studying antibiotic exposure by month, there was no difference in the rates of AD in infants who were or were not exposed to antibiotics during months of life 2 through 12 months. Of infants who received antibiotics at week 2 of life, 20.2% developed eczema, compared with 26.9% who did not receive antibiotics (P = .016).

Antibiotic exposure during the first month of life, especially at week 2 of life, may be associated with decreased risk of AD development. However, there was a modest, but statistically significant difference in AD prevalence between infants who received antibiotics (22.3%) compared with infants who did not during their first year of life (28%). Prenatal antibiotic exposure was not associated with a difference in the development of AD.

This retrospective study looked closely at antibiotic use prenatally and the immediate postnatal age group at weekly intervals to better identify a possible “critical window” of antibiotic use that affects the development of AD. It had over 4000 participants that included a representation of underserved populations.

Antibiotic use, particularly in the first month of life, appeared to have a protective effect on the development of AD. The study was limited by its retrospective nature and it being conducted at a single center. Additionally, antibiotic use outside of the medical center and AD diagnoses that were not captured by International Classification of Diseases codes may have been missed. Multicenter prospective studies are warranted to further review the relationship between early antibiotic use and its effect on the development of AD.