Oral aphthosis in children: baseline clinical features associated with diagnostic categories in a single-centre retrospective cohort.

UNLABELLED: We examined associations between clinical features at first paediatric rheumatology assessment and eventual benign/self-limited (Group 1) versus systemic/chronic (Group 2) diagnostic categories in children referred for recurrent oral aphthosis. We retrospectively studied 404 children evaluated from 2010 to 2024. Baseline clinical variables were documented at the first rheumatology visit before final diagnosis; follow-up assessments were performed every 3-6 months. Eight prespecified clinical features entered multivariable logistic regression. Firth and diagnosis-exclusion sensitivity analyses addressed sparse data and diagnostic-composition bias. CRP/ESR and HLA-B51 were analysed separately. The cohort included 198 males (49.0%) and 206 females (51.0%); age at symptom onset ranged from 2 months to 15 years. Group 1 included 267 patients, and Group 2 included 137. Recurrent fever (aOR, 0.25; 95% CI, 0.12-0.49) and tonsillitis/pharyngitis (aOR, 0.17; 95% CI, 0.09-0.32) were associated with lower odds of Group 2, whereas genital aphthosis was associated with higher odds (aOR, 13.08; 95% CI, 3.04-56.32). After simultaneous exclusion of PFAPA and Behçet disease spectrum diagnoses, the fever/pharyngitis associations were attenuated, and the genital aphthosis estimate became uninformative because of sparse data, while arthralgia/arthritis remained positively associated. Elevated CRP/ESR was not independently associated with Group 2; HLA-B51 positivity was more frequent in Group 2 among selectively tested patients. CONCLUSION: In children referred for recurrent oral aphthosis, recurrent fever and pharyngitis were mainly associated with benign/self-limited diagnoses, whereas genital aphthosis was mainly associated with systemic/chronic diagnoses. These associations were largely driven by PFAPA and Behçet spectrum disease, respectively. Articular involvement showed the most consistent positive association across sensitivity analyses. These findings are descriptive, not a validated diagnostic prediction tool. WHAT IS KNOWN: • Oral aphthae are common and non-specific in childhood. • Associated symptoms guide diagnostic work-up and follow-up. WHAT IS NEW: • PFAPA-characteristic features clustered with benign/self-limited diagnoses; genital aphthosis with systemic/chronic diagnoses. • After PFAPA/Behçet spectrum exclusion, these signals attenuated; articular involvement was the most consistent positive association.