Early-life factors may flag infants at risk of atopic dermatitis

Researchers in Thailand have developed prediction models that use information gathered in early infancy to estimate which babies are likely to develop atopic dermatitis within their first two years of life. The study, a prospective cohort analysis published in Allergy, Asthma & Clinical Immunology, found that higher household income and a family history of allergy were associated with an increased risk of early-onset atopic dermatitis, while regular moisturizer use was associated with a reduced risk.

The team enrolled healthy infants aged 42 to 90 days at Songklanagarind Hospital between August 2020 and October 2021. Demographic and clinical characteristics were recorded at baseline, and atopic dermatitis was assessed at 12 and 24 months of age using the Hanifin and Rajka diagnostic criteria. Of the 671 infants included in the analysis, 55, or 8.2 percent, developed atopic dermatitis within the first two years of life.

To build the models, the researchers used multivariable logistic regression with two different variable selection approaches: stepwise selection and the least absolute shrinkage and selection operator, known as LASSO. Performance was evaluated using sensitivity, specificity, the area under the receiver operating characteristic curve, F1 score, and calibration plots.

The two approaches produced different trade-offs. The stepwise-selected model showed lower sensitivity than the LASSO-selected model, 68 percent versus 74 percent, meaning it missed a somewhat larger share of infants who went on to develop the condition. However, the stepwise model was more specific, 75 percent versus 66 percent, and achieved a higher area under the curve, 74 percent versus 67 percent. Both models were similar in F1 score, at 26 percent, and showed similar calibration.

The authors concluded that prediction models based on early-life clinical factors demonstrated promising performance in identifying infants at risk of atopic dermatitis, and that such models may support early risk stratification and prevention-oriented counseling in infancy. The reported associations, including the link between regular moisturizer use and lower risk, come from an observational cohort and do not by themselves establish that moisturizer use prevents the disease.

The study is a secondary analysis of data from participants enrolled in the V114-032 (PNEU-ERA) clinical trial, which was prospectively registered at ClinicalTrials.gov as NCT04193215. The cohort analysis of atopic dermatitis prediction did not involve a separate interventional protocol.

Several caveats apply to the findings. The cohort was drawn from a single hospital, the number of atopic dermatitis cases was modest at 55, and the F1 score of 26 percent indicates that both models had limited overall balance between correctly identifying cases and avoiding false positives. The article was published as a peer-reviewed accepted manuscript shared early to provide faster access, and it is subject to further edits before the final version of record.

Atopic dermatitis is the most common chronic inflammatory skin disease in childhood, according to the study’s background. The authors note that early identification of high-risk infants may support preventive interventions and reduce the subsequent burden of allergic diseases, making accurate early risk stratification a meaningful goal for pediatric care.

Subject of Research: Prediction models for early-onset atopic dermatitis in infants using early-life clinical factors

Article Title: Prediction models for early-onset atopic dermatitis in infancy: a prospective cohort study

Article References: Eakpanit, E., Laoprasopwattan, K., Sangsupawanich, P., Tassanakijpanich, N., Chumchuen, K., & Koosakulchai, V. (2026). Prediction models for early-onset atopic dermatitis in infancy: a prospective cohort study. Allergy, Asthma & Clinical Immunology. https://doi.org/10.1186/s13223-026-01068-4

Image Credits: AI Generated

DOI: 10.1186/s13223-026-01068-4

Keywords: atopic dermatitis, infants, prediction models, prospective cohort study, LASSO, logistic regression, risk stratification, moisturizer use, family history of allergy, pediatric allergy, Songklanagarind Hospital, Hanifin and Rajka criteria