Celiac disease (CD), also known as celiac sprue or gluten-sensitive entheropathy, can be defined as a chronic inflammatory intestinal disease characterised by nutrient malabsorption and improvement after the withdrawal of gluten (found in wheat, barley) from the diet. Prevalence of CD ranges from 1:85 to 1:300 have been reported for CD in Western countries 123456. In addition to the classical gastrointestinal presentation (diarrhoea, abdominal distension, vomiting, weight loss and pallor) CD can cause minimal intestinal damage and weak or absent systemic symptomatology (also known as "silent form"). In these patients the lack of symptoms can persist for a long time, while the biopsy of the bowel shows the typical atrophy of intestinal mucosa 7. It is also well recognized the association of CD with several complications, as lymphomas, autoimmune and degenerative nervous system diseases 8910.

The oral cavity, a part of gastrointestinal system 11, can also be affected by several abnormalities in patients with CD. As the mouth is very easy to examine, oral lesions can provide a valuable clinical clue for early diagnosis of CD 12; in fact among the atypical aspects of CD (extra-intestinals), in the international literature has been reported some affections interesting the oral cavity, the most common are recurrent aphthous stomatitis (RAS) 131415 and dental enamel defects 813161718192021, in addition have been described the association between CD and unspecific forms of atrophic glossitis 22, oral manifestations of dermatitis herpetiformis 23, Sjögren's syndrome 2425 and oral lichen planus 2627. These disorders, in absence of a typical intestinal symptomatology, can represent useful clues for a timely diagnosis 722.

However, data from literature are often controversial, probably because of different geographical origin of patients studied and lack of adequate controls. Finally, no studies have been performed, in CD patients of a Central Region of Italy (Ancona, Marche, Italy)

The aim of this case-control study was to assess prevalence of dental hard and oral soft tissues changes generally considered celiac-related (e.g. RAS, enamel hypoplasia, dermatitis herpetiformis and atrophic glossitis) and to verify if cases are more likely to be affected by any of the oral diseases considered.

Fifty CD patients, aged between 3 and 18 years old and living in the Region of Marche, were enrolled in the study. CD was diagnosed at Paediatric Department of the University Politecnica of Marche (Ancona, Italy), and the diagnosis of CD was based on serological tests (Ab-htTG IgA, Ab-htTG IgG, AGA IgA, AGA IgG, EMA IgA, EMA IgG), small-bowel biopsy during esophago-gastro-duodenoscopy (EGDS) and histological evidence of villous atrophy with crypt hyperplasia and increase in intraepithelial lymphocytes (normal, 10–30 per 100 epithelial cells), and the disappearance of the symptoms and normalization of serum anti-tTG and/or EMA after gluten-free-diet (GFD) 2829. The control group was recruited by simple randomization at a Primary and Secondary Public School of Ancona, during an healthy prevention programme for oral disease, matched one-to-one and without any significant differences with study group for geographical area, age and gender (p > 0.2 by t-Student and chi-square test, respectively). These young individuals neither reported any gastrointestinal diseases and not have a family history of CD.

Patients were examined for hard tissue changes (i.e. dental enamel defects) and soft tissue lesions (RAS, dermatitis herpetiformis and atrophic glossitis). Patients with CD and healthy individuals were examined by a single observer. Informed consent was obtained by parents who were also asked about previous episodes of RAS affecting child/children.

The enamel defects affecting deciduous and permanent teeth were graded 0 to IV according to Aine's classification 17 with a special attention to symmetric anomalies.

Soft tissues examination was carried out with conventional dental chairs, artificial light, flat mirrors, monouse probe and sterile gauzes.

With regard RAS, we registered both lesions clinically observed and ulcerative events referred by parents or reported by hospital clinical records. They were classified into minor, major and herpetic aphthous ulcers 30, according to dimension, form, localization and evolutionary tendency, and also rate of occurrence was registered. Atrophic glossitis was diagnosed on the basis of clinical features and oral mucosal lesions due to dermatitis herpetiformis were assessed by both clinical features and histological/immunofluorescence studies.

Enamel alterations were observed in 13/50 (26%) subjects with CD and in 8/50 (16%) controls, with a ratio male-female of 1:2 for the celiac group and 2:1 for control group (p > 0.2; OR = 1.8446; 95% CI = 0.6886: 4.9414). With respect to the severity score of hypoplasia, 10/13 CD patients showed lesions of degree 1 and 3/13 degree 2, in controls all were in degree 1. The grade 1 enamel defects were generally localized on incisor surfaces (for the anterior sectors) (Figure 1) and cuspid surfaces (for the posterior sectors), with dimensions from 1 to 3 mm and with a round-oval form, while that of grade 2 were on the canine and premolar vestibular surface. The colour alterations were white-yellowish, with clear margins, opaque and smooth surface.

Episodes of RAS occurred in 36% of CD patients (18/50) vs 12% of controls (6/50) (p = 0.0091; OR = 4.125; 95% CI = 1.4725: 11.552) with a male-female ratio of 1:1 and 2:3, respectively (Figure 2). In CD patients RAS showed greater rate of recurrence than in controls. Atrophic glossitis was reported in 4 cases and one control, and dermatitis herpetiformis in one patient with CD and none of subjects without CD.

Recent epidemiology data showed the prevalence of CD to approach 1% of the general population 31323334. However, the clinical presentation of CD seems to differ from the typical form observed in past years, as almost 50% of the patients with newly diagnosed CD do not present with gastrointestinal symptoms 3536. Thus, in order to identify the greatest number of "atypical" or "silent" CD patients and prevent long-term complications, it has been suggested that the clinicians should investigate those subjects who present "indirect" signs of CD, such as chronic anaemia 37, hyper-transaminasemia or hyperamylasemia of unknown origin 3839, osteoporosis 40, autoimmune thyroid disorders 41.

As abnormalities of the oral cavity have been reported in CD, non-invasive clinical examination of the oral cavity can contribute to identify patients with atypical or silent CD 1314171842.

As regards to changes of dental tissues, we did not found CD patients more likely to suffer from systematic and symmetric enamel defects. Indeed, a wide range of frequencies of enamel defects in CD patients has been reported in other studies 17434445464748; our data are in agreement with other studies performed in Italy (Table 1) and the high frequency of enamel defect found in controls, as well as its severity, is likely to be related to environmental, dietetic and genetic factors 46. Further studies are warranted to clarify the pathogenesis of this defect as nutritional, immunologic or genetic factors (association with the HLA DR3 allele) has been hypothesized 4549. With regard to celiac patients, enamel defects have been correlated to an altered phosphate-calcium metabolism and/or formation of antibodies against the matrix of enamel organ. The antigen correlated to class II molecules of the MHC could prime an immunity movement against the enamel organ, from which a mineralization disorder could derive 18. In addiction, there is no strong evidence that these anomalies are correlated with the nutritional status, vitamin D deficiency or to an excess of fluoride incorporation. Current evidence suggests that an autoimmune pathogenesis is more likely, as enamel defects are also present in autoimmune diseases, such as some polyendocrine syndromes 46.

With respect to oral soft lesions, we confirmed that CD patients are likely to suffer from RAS compared with healthy controls, especially before the gluten-free diet.

In our celiac population RAS was found in 26 % of CD patients with an OR of 4.12 in comparison with the controls. Even if a wide range of frequencies have been reported (Table 2) our data show the highest prevalence of RAS with respect to other Italian studies.

In agreement with Sedghizadeh et al. 14, we suggested to consider RAS as a "risk indicator" of CD more than CD as a risk factor for RAS, although no definitive statement is possible on their predictive role for CD.

In addition the term "recurrent aphthous stomatitis" should be reserved to recurrent oral ulcer that present in patients without systemic diseases, while ulcers that have a clinical appearance similar to RAS, but found in patients with systemic disorders (such as CD) should be termed "aphthous-like ulcers" 50. Even if the diagnostic criteria of RAS used in this study (namely, medical history and/or presence of detectable lesions) may represent a major limitation of present research, it is well accepted that recurrent and episodic nature of oral ulcerations requires medical history to be an important part of the diagnostic process.

RAS is often associated to haematinic (iron, folate, vitamin B12) deficiency 5152; since atypical or latent CD may not manifest itself with gastrointestinal signs/symptoms but often with iron/folate deficiency 53545556 we suggest that when patients show persistent RAS they should be examined for haematinic deficiencies. Only if one or more of these deficiencies are present, they should be screened for CD.

In conclusion, our data from central Italy confirming the higher prevalence of RAS or aphthous-like ulcers in patients with CD validate the hypothesis of their pathogenetic predisposition to oral mucosal lesions more than hard dental tissue lesions; further investigations are warranted to clarify the predictive role of these lesions in screening oligosymptomatic or asymptomatic CD.