IMPORTANCE

Atopic dermatitis (AD) is the most common inflammatory disease in childhood, and children with AD are more likely to develop other allergic diseases, including food allergy, allergic rhinitis, and asthma.

OBJECTIVE

To determine the phenotypes of AD expression across 12 US birth cohorts and identify factors associated with phenotype and development of allergic diseases.

DESIGN, SETTING, AND PARTICIPANTS: This cohort study compiled longitudinal data from 12 observational US birth cohorts across decades (children born from April 1980 to June 2019) in the Environmental Influences on Child Health Outcomes (ECHO) Children's Respiratory and Environmental Workgroup with follow-up to September 2022. Participants were enrolled prenatally; children with 3 or more AD assessments across the first 84 months of life were included in analyses. Data were analyzed from December 2020 to April 2024.

EXPOSURES

Exposures included decade of birth, cohort type (population-based or high-risk), family history of asthma (mother, father, or sibling), birth order, gestational age at birth, delivery mode, breastfeeding, pet exposure, antibiotic use, environmental tobacco smoke exposure, allergic sensitization, peripheral blood eosinophil count, and total IgE.

MAIN OUTCOMES AND MEASURES

Primary outcomes were AD phenotype, food allergy, allergic rhinitis, asthma, and wheeze. Longitudinal latent class analysis was used to identify underlying longitudinal patterns of AD expression, and associations of AD phenotype with allergic outcomes were examined using logistic regression, multinomial logistic regression, and linear regression.

RESULTS

In 5314 children from 9 cohorts (1896 born in the 2000s [35.7%]; 2585 female [48.6%]; 1083 Black or African American [20.4%]; 3344 White [62.9%]; 350 other reported race [6.6%; including 8 American Indian or Alaska Native (0.2%); 58 Asian (1.1%); 4 Native Hawaiian or Pacific Islander (0.1%) and 280 multiracial or with any race not otherwise specified (5.3%)]), 3382 (63.6%) were from a population-based cohort, while 1932 (36.4%) were from a high-risk cohort. AD prevalence ranged from 24.1% (540 children) to 28.4% (1156 children) at each time point, and 5 phenotypes of AD were identified: transient early AD, early AD with potential reoccurrence, late-onset AD, persistent AD, and minimal or no AD. Compared with White children, Black children were at higher risk for AD (transient early AD: aOR, 3.26; 95% CI, 2.06-5.18; early AD with potential reoccurrence: aOR, 3.72; 95% CI, 2.35-5.90; persistent AD: aOR, 2.01; 95% CI, 1.54-2.63), as were children with other reported race (transient early AD: aOR, 2.31; 95% CI, 1.13-4.70; early AD with potential reoccurrence: aOR, 3.27; 95% CI, 1.73-6.18). Female children were significantly less likely to have early AD with potential reoccurrence (aOR, 0.45; 95% CI, 0.27-0.74) and persistent AD (aOR, 0.60; 95% CI, 0.49-0.74) than male children. Compared with miniml or no AD, phenotypes with early AD expression were associated with food allergy (transient early AD: adjusted odds ratio [aOR], 2.15; 95% CI, 1.48-3.08; early AD with potential reoccurrence: aOR, 2.43; 95% CI, 1.66-3.50; persistent AD: aOR, 2.26; 95% CI, 1.84-2.78), later AD expression was associated with allergic rhinitis (late-onset AD: aOR, 1.84; 95% CI, 1.38-2.43; persistent AD: aOR, 2.02; 95% CI, 1.64-2.48), and any AD disease was associated with asthma.

CONCLUSIONS AND RELEVANCE

In this birth cohort study of 5314 children, timing of AD expression was associated with increased risk for atopic march pathways. Identifying risk factors for AD phenotypes may inform targeted therapeutic prevention strategies.