부산대학교 도서관

Background: Skin barrier dysfunction is associated with the development of atopic dermatitis (AD), however methods to assess skin barrier function are limited. We investigated the use of electrical impedance spectroscopy (EIS) to detect skin barrier dysfunction in children with AD of the CARE (Childhood AlleRgy, nutrition, and Environment) cohort. Methods: EIS measurements taken at multiple time points from 4 months to 3‐year‐old children, who developed AD (n = 66) and those who did not (n = 49) were investigated. Using only the EIS measurement and the AD status, we developed a machine learning algorithm that produces a score (EIS/AD score) which reflects the probability that a given measurement is from a child with active AD. We investigated the diagnostic ability of this score and its association with clinical characteristics and age. Results: Based on the EIS/AD score, the EIS algorithm was able to clearly discriminate between healthy skin and clinically unaffected skin of children with active AD (area under the curve 0.92, 95% CI 0.85–0.99). It was also able to detect a difference between healthy skin and AD skin when the child did not have active AD. There was no clear association between the EIS/AD score and the severity of AD or sensitisation to the tested allergens. The performance of the algorithm was not affected by age. Conclusions: This study shows that EIS can detect skin barrier dysfunction and differentiate skin of children with AD from healthy skin and suggests that EIS may have the ability to predict future AD development. [ABSTRACT FROM AUTHOR]