Risk of Developing Pediatric Uveitis Among Patients With Early-Onset Atopic Dermatitis.

Importance: The relationship between atopic dermatitis (AD) and pediatric uveitis may be underexplored, warranting large-scale, multicenter studies.

Objective: To evaluate the risk of pediatric uveitis among children with early-onset AD compared with a matched control population.

Design, setting, and participants: This cohort study used aggregated electronic health records of US patients with early-onset AD and matched controls from January 1, 2004, through December 14, 2024, sourced from health care organizations in the collaborative research network TriNetX. Patients with early-onset AD and matched controls without AD were included in the analysis; those with uveitis prior to AD diagnosis were excluded. Propensity score matching was applied to balance baseline characteristics. The analyses were conducted on December 14, 2024.

Exposure: International Classification of Diseases, 10th Revision (ICD-10) diagnosis code for AD.

Main outcomes and measures: The primary outcome was the hazard ratio (HR) for developing pediatric uveitis in the AD cohort compared with the matched controls. Cox proportional hazards models were applied to assess the risk.

Results: A total of 114 889 patients were identified in the AD cohort (mean [SD] follow-up, 6.0 [3.3] years; mean [SD] age, 0.5 [0.7] years; 64 817 male [56.4%]) and the control cohort (mean [SD] follow-up, 6.6 [3.7] years; mean [SD] age, 0.6 [0.8] years; 65 377 male [56.9%]) after matching. The AD cohort demonstrated a higher risk of developing pediatric uveitis compared with controls (94 [0.08%] vs 58 [0.05%]; HR, 1.92 [95% CI, 1.38-2.66]). Sensitivity analyses among patients without dupilumab use (89 of 113 284 [0.08%] vs 59 of 113 284 [0.05%]; HR, 1.77 [95% CI, 1.27-2.46]) and those without autoimmune conditions (80 of 114 425 [0.07%] vs 61 of 114 425 [0.05%]; HR, 1.52 [95% CI, 1.09-2.12]) similarly indicated an increased risk in the AD cohort. Additionally, patients with severe AD had a higher risk of developing pediatric uveitis compared with those with nonsevere AD (12 of 3004 [0.40%] vs 97 of 126 482 [0.08%]; HR, 3.64 [95% CI, 2.00-6.66]).

Conclusions and relevance: This cohort study of children with early-onset AD found an elevated risk of pediatric uveitis compared with matched controls, independent of autoimmune conditions or dupilumab use. These findings support the potential need to consider ophthalmologic monitoring in children with early-onset AD to try to detect and subsequently manage uveitis if it develops.