Phenotypes of atopic dermatitis identified by cluster analysis in early childhood

• Published in: Journal of dermatology Vol. 46; no. 2; pp. 117 - 123
• Main Authors: Seo, Euri, Yoon, Jisun, Jung, Sungsu, Lee, Jina, Lee, Beom Hee, Yu, Jinho
• Format: Journal Article
• Language: English
• Published: England Wiley Subscription Services, Inc 01.02.2019
• Subjects: Age; Allergens; Atopic dermatitis; Blood; child, preschool; Children; Cluster analysis; Dermatitis; Eczema; Food allergies; Immunoglobulin E; Immunoglobulins; infant; Leukocytes (eosinophilic); phenotype; Phenotypes; Skin diseases; Water loss; phenotype; atopic dermatitis; cluster analysis; infant; child, preschool
• ISSN: 0385-2407; 1346-8138
• DOI: 10.1111/1346-8138.14714

Atopic dermatitis is a chronic, relapsing, inflammatory skin disease that usually appears in early childhood and develops into a heterogeneous disease during childhood. The clinical course and treatment for atopic dermatitis can differ according to its phenotype and/or endotype. This study aimed to identify clinical phenotypes of atopic dermatitis in early childhood. Data were obtained from 572 children under 3 years of age with atopic dermatitis. Cluster analysis applied to 11 variables, and we identified four clusters of atopic dermatitis. Children in cluster A (n = 141) had early‐onset atopic dermatitis with high blood eosinophil counts, serum total immunoglobulin E and rates of sensitization to food allergens. Children in cluster B (n = 218) had early‐onset atopic dermatitis with low blood eosinophil counts, serum total immunoglobulin E and rates of sensitization to both food and inhalant allergens. Children in cluster C (n = 53) had early‐onset atopic dermatitis with high C‐reactive protein levels and white blood cell counts. Children in cluster D (n = 160) had middle‐onset atopic dermatitis with high serum total immunoglobulin E and rates of sensitization to inhalant allergens. Cluster A had the highest Scoring for Atopic Dermatitis and transepidermal water loss values. Age at onset, age at diagnosis, white blood cell count, eosinophil count, C‐reactive protein and serum total immunoglobulin E level were the strongest predictors of cluster assignment. Analysis of these six variables alone resulted in correct classification of 95.5% of the subjects. These results support the heterogeneity of atopic dermatitis, even in early childhood.