Sociodemographic and environmental profiles associated with atopic dermatitis severity among children

Summary: Abstract Body: Racial and ethnic disparities in atopic dermatitis (AD) severity exist among children in the United States. The combinations of social and environmental factors contributing to these disparities remain unclear. We aimed to identify population groups and their characteristics (profiles) that are associated with pediatric AD severity. We performed a cross-sectional analysis of children enrolled in the Pediatric Eczema Elective Registry and their associated zip code-based social determinants of health (SDoH) and air pollution (particulate matter [PM] 2.5) data from the American Community Survey and Environmental Protection Agency, respectively. The primary outcome was AD control (poor vs. good). Latent class analysis was used to identify subgroups (latent classes) based on individual and zip code-level demographic, clinical, SDoH and PM2.5 data. Logistic regression evaluated the associations between latent classes and AD control. In total, 5,212 children were included in the study. Mean (standard deviation) age was 7.4 (4.1) years; 52.7% were female; racial/ethnic distribution was 34.4% White, 55.4% Black, 10.2% Hispanic. Six latent classes were identified: Class 1 (socioeconomically advantaged), Class 2 (minoritized and socioeconomically disadvantaged, housing challenged, higher PM2.5 exposure, low atopy); Class 3 (minoritized and socioeconomically disadvantaged, low PM2.5 exposure); Class 4 (multiracial, middle class, immigrant population); Class 5 (White population, lower education, low PM2.5 exposure); Class 6 (minoritized and socioeconomically disadvantaged, higher PM2.5 exposure). Compared to Class 1, Classes 2 and 6 had higher odds of poor AD control (odds ratios [OR] 1.34 and 1.27; 95% confidence intervals [CI] 1.08–1.67 and 1.08–1.50, respectively). Classes 3 and 5 had lower odds of poor AD control (OR 0.62 [CI 0.49–0.77] and OR 0.50 [CI 0.41–0.62], respectively). Our study identifies specific exposure profiles, especially PM2.5 levels, associated with AD severity that can guide interventions to address disparities in pediatric AD outcomes. Robert Fitzsimmons¹, Ole Hoffstad¹, David J. Margolis¹, Daniel Shin¹, Junko Takeshita¹ 1. University of Pennsylvania, Philadelphia, PA, United States. Minoritized Populations and Health Disparities Research