Evaluation of Childhood Atopic Dermatitis and Environmental Factors in Turkey with Decision Tree Model.

OBJECTIVE: This study aims to examine the relationship between atopic dermatitis (AD), one of the most common dermatological conditions in children, and environmental factors, including meteorological variables and air pollution. METHODS: This retrospective cross-sectional study analyzed the medical records of 21,407 pediatric patients aged 0 to 18 years who presented to the city hospital in Agri, Turkey, between 2020 and 2024. Admission dates were matched with meteorological data (wind speed, atmospheric pressure, humidity, temperature) and air pollution indicators (PM10, SO2, NO2, NOx, NO, O3). Statistical analyses included t-tests, correlation analyses, binary logistic regression, and a CHAID decision tree model. RESULTS: AD accounted for 10.1% of all dermatology-related visits. AD admissions increased particularly during the first half of the year and were significantly associated with higher O3 levels, whereas increased PM10 levels were associated with a lower likelihood of AD admissions. Logistic regression showed that age, sex, semiannual period, atmospheric pressure, PM10, and O3 were significant predictors of AD. The decision tree model identified age, period, and O3 as the strongest discriminating variables for AD. CONCLUSION: AD was found to be more sensitive to environmental and seasonal variations compared with other dermatitis types. In particular, elevated ozone levels and temporal factors played a notable role in increasing AD presentations. These findings may inform environmental risk management and preventive strategies for children with AD.