Background: Atopic dermatitis (AD) is a prevalent disease with variable natural history. Longitudinal birth cohort studies provide an opportunity to define subgroups based on disease trajectories, which may represent different genetic and environmental pathomechanisms.
Objective: To investigate the existence of distinct longitudinal phenotypes of AD and test whether these findings are reproducible in two independent cohorts.
Methods: The presence of AD was examined in two birth cohort studies including 9,894 children from the UK (ALSPAC) and 3,652 from the Netherlands (PIAMA). AD was defined by parental report of a typical itchy and/or flexural rash. Longitudinal latent class analysis was used to investigate patterns of AD from birth to the age of 11 to 16 years. We investigated associations with known AD risk factors, including FLG null mutations, 23 other established AD-genetic risk variants and atopic comorbidity.
Results: Six latent classes were identified, representing subphenotypes of AD, with remarkable consistency between the two cohorts. The most prevalent class was early-onset-early-resolving AD, which was associated with male gender. Two classes of persistent disease were identified (early-onset-persistent and early-onset-late-resolving); these were most strongly associated with the AD-genetic risk score as well as personal and parental history of atopic disease. A yet unrecognised class of mid-onset-resolving AD, not associated with FLG mutations, but strongly associated with asthma, was identified.
Conclusion: Six classes based on temporal trajectories of rash were consistently identified in two population-based cohorts. The differing risk factor profiles and diverse prognoses demonstrate the potential importance of a stratified medicine approach for AD.
Clinical Implications: Atopic dermatitis ranges from a transient condition to lifelong morbidity. This study has identified distinct subphenotypes of atopic dermatitis in children, which could indicate the importance of a stratified approach to management of this complex disease.
- Atopic dermatitis
- Latent class analysis