Child's Health · Published 2025-12-08 · DOI 10.22141/2224-0551.20.8.2025.1928
Background. The high prevalence of asymptomatic sensitization complicates food allergy diagnosis, leading to unnecessary elimination diets and overuse of invasive oral food challenges. There is a critical need to develop objective predictive scoring models, particularly for common yet less studied allergens such as salmon (Sal s 1). The purpose was to identify the main clinical-anamnestic and immunological factors for building a logistic regression predictive model of the risk of developing clinical manifestations in sensitization to the β-parvalbumin allergen of salmon (Sal s 1). Materials and methods. We examined 239 children aged 1 to 18 years with confirmed sensitization to salmon β-parvalbumin (Sal s 1). The patients’ clinical-anamnestic data and immunological test results were analyzed; using a logistic regression model, the risk of developing clinical manifestations of food allergy was assessed. A preliminary diagnosis of food allergy was established based on the EAACI 2023 guidelines. Results. Among the 239 patients sensitized to Sal s 1, clinical manifestations of food allergy were detected in 144 cases (60.3 %). Binary logistic regression, which included 10 significant factors after refinement (AUC 0.83), determined that the strongest positive predictors of clinical manifestation risk were: the number of β-parvalbumin molecules to which the patient was sensitive in the sensitization profile (OR 1.99), the presence of respiratory allergy (OR 1.84), sensitization class 4 (OR 1.85), and atopic dermatitis (OR 1.59). The strongest protective effect was found for sensitization class 1 (OR 0.26). Conclusions. The developed logistic regression predictive model (AUC 0.83) allows for the assessment of individual risk for clinical manifestations of food allergy in children sensitized to salmon β-parvalbumin. Key prognostic factors include age, the degree of Sal s 1 sensitization, sensitivity to other β-parvalbumins, the presence of atopic dermatitis, and respiratory allergy. This model can be used for non-invasive stratification of patients based on the risk level of developing clinical manifestations of food allergy, allowing for the implementation of elimination measures without risky provocative challenges in high-risk children.
Abstract from DOAJ. Public domain (CC0 1.0).
Read the article at the publisher →