In this study of a large group of patients with RAS, a 2.83% prevalence of GSE was observed, as compared with an estimated prevalence of 0.9% in the general population of Iran .
In order to avoid the controversy in the definition of CD, we used the term "gluten sensitive enteropathy" rather than celiac disease to describe patients with any degrees of intestinal damage together with positive serologic tests.
It has been reported that RAS is at least among the fifth commonest presentations of CD [14–16]. Furthermore, oral mucosal lesions or dental enamel defects may be the sole presenting features of celiac disease .
Despite detailed investigations, RAS still has an unknown etiology and poorly effective management [18, 19]. Genetic, immunological and microbial factors may play a role in the pathogenesis of RAS, whereas attacks may be precipitated by local trauma, stress, food intake, some drugs, hormonal changes or vitamin and trace element deficiencies . RAS can arise in some systemic disorders including: Behcet's disease [20, 21], Sweet's syndrome [22, 23], MAGIC syndrome , inflammatory bowel disease [25, 26] and gluten-sensitivity enteropathy (celiac disease) .
The association between RAS and gluten-sensitivity enteropathy (GSE) was proposed in 1976 by Ferguson et al  when they found 24% of patients with RAS showed histological evidence of CD on jejunal biopsy. Nevertheless, there is still considerable dispute concerning the actual prevalence of CD among patients with RAS, as different studies have reported different prevalence of CD in RAS patients [29–34]. On the other hand in recent years, some articles are published which expressed little or no significant etiological link between RAS and CD, and added that screening RAS patients for key serological markers of CD is of little clinical value [10, 1]. Currently, there is no approved recommendation which can be used by clinicians to approach patients with RAS regarding celiac disease. Comparing to the previous studies, this study with a large group of patients and by using two more sensitive and specific serologic tests plus duodenal biopsy helped us reach a reliable conclusion. Prevalence of GSE in patients with RAS was 2.83%, which is about 3-fold higher than that expected in general population of Iran (0.9%) .
The effect of gluten-free diet (GFD) on remission of RAS is still uncertain, as dietary withdrawal of gluten occasionally results in significant benefit whereas some studies reported it ineffective [11, 35–36]. Four patients accepted to start GFD, and all of them showed a significant improvement within 2–6 months after beginning of GFD. Furthermore anemia resolved after 6 months of follow up in the two patients who suffered from anemia.
Many physicians may still consider the gastrointestinal signs and symptoms as a main manifestation of celiac patients whereas recent studies demonstrated that gastrointestinal presentations may be absent in GSE patients especially in the beginning of the disease. In this study, none of our GSE patients had any gastrointestinal symptoms. Therefore, gastrointestinal symptoms are sometimes absent in the setting of the disease and RAS could be the first or the sole presentation of GSE.
Our study has some limitations. We did not take duodenal biopsies from the patients who had negative serological tests. It has been reported that the sensitivities of the serological tests are decreased in GSE patients with minor mucosal damages [37, 38]. We cannot exclude the possibility of missing some GSE patients with negative serological tests and Marsh I/II mucosal lesions (e.g. seronegative GSE). However, a patient with negative serological test and duodenal mucosal lesion may suffer from other disorders like autoimmune enteropathy, giardiasis, common variable immunodeficiency, tropical sprue, peptic duodenitis, Crohn's disease etc. Including such patients (e.g. those with negative serological tests with duodenal mucosal damage) in the spectrum of GSE could increase the rate of false positive results; unless symptomatic and histological improvements are confirmed by gluten free diet. Therefore, in the epidemiological studies, a positive result from a highly specific serological test (e.g. EMA, or tTG) together with any degree of duodenal mucosal lesion provide reasonable criteria for identifying patients with GSE.