Ethnic Differences in Clinical Manifestations of Behçet's Disease: A Retrospective Cohort Study of Turkic and Non‐Turkic Patients
Shokufeh Sadeghi, Seyedeh Tahereh Faezi, Mohsen Rokni, Elham Madreseh, Farhad Shahram, Majid Alikhani, Somayeh Motamed, Fereydoun DavatchiABSTRACT
Introduction
Behçet's disease (BD) is a multisystem inflammatory vasculitis with clinical heterogeneity that may vary across ethnic groups within the same country. This study aimed to evaluate the influence of ethnicity on the clinical manifestations, immunogenic laboratory findings and disease course of BD by comparing Turkic and non‐Turkic patients in a large Iranian referral cohort.
Methods
In this retrospective cohort study, 2697 patients with BD followed at the BD clinic of Shariati Hospital in Tehran, Iran, from 2007 to 2019 were included.
Results
Patients were stratified by ethnicity into Turkic ( N = 1037; 38.5%) and non‐Turkic ( N = 1660; 61.5% [with Persian/Fars, N = 980 subgrouping]) groups, and demographic, clinical manifestations, and immunogenic laboratory findings were extracted from medical records. Mean age at BD onset was 25.72 ± 10.02 years in Turkic patients and 24.69 ± 10.23 years in non‐Turkic patients, and males comprised 61.3% and 57.1% of the two groups, respectively. Human leukocyte antigen (HLA)‐B5 positivity was 55.1% in Turkic versus 53.1% in non‐Turkic patients, and HLA‐B51 positivity was 48.7% versus 46.5%, respectively. Most clinical manifestations showed similar prevalence across groups, including oral aphthous ulcers (98.7% vs. 99.0%), genital aphthous ulcers (60.5% vs. 62.0%), uveitis (50.1% vs. 53.2%), and deep vein thrombosis (5.3% vs. 5.1%). Superficial phlebitis was more frequent in Turkic patients ( p = 0.02), and oral aphthous ulcers occurred earlier in Turkic patients ( p = 0.023). A three‐class latent class model demonstrated the best fit (BIC = 14,986.58). In multinomial logistic regression analysis, Turkic patients had slightly higher odds ratio of the ocular‐predominant phenotype (OR = 1.10, 95% CI: 0.91–1.33; p = 0.320) and slightly lower odds ratio of the multisystem phenotype (OR = 0.94, 95% CI: 0.68–1.30; p = 0.692); however, neither association was statistically significant. Female gender was associated with higher odds ratio of the ocular‐predominant phenotype than the mucocutaneous‐predominant phenotype (OR = 1.38, 95% CI: 1.14–1.67; p = 0.001).
Conclusion
In this large Iranian BD cohort, Turkic and non‐Turkic patients exhibited broadly similar disease phenotypes and comparable HLA‐B51 frequencies, with small but potentially relevant ethnic differences in the timing of oral ulcers and the frequency of superficial phlebitis. These findings suggest that ethnicity may influence BD phenotype through mechanisms beyond HLA‐B51, possibly involving other immune‐related genetic variants such as tumor necrosis factor (TNF)‐α, endoplasmic reticulum aminopeptidase 1(ERAP1), Interleukin (IL)‐10. Of note, the significant differences observed in Turkic versus non‐Turkic comparisons were not present in the Turkic versus Persian (Fars) comparison, indicating that the findings are not driven by a single ethnic subgroup.