Abstract
Objectives
The aim of this systematic review was to compare the presence of enamel defects and aphthous stomatitis between celiac patients and healthy controls.
Data sources
A systematic review of articles selected from MEDLINE, EMBASE and Google Scholar was performed by two independent operators. Additional studies hand-searched and found in the principal dental and gastroenterology journals were included.
Study selection
Only controlled studies on celiac patients compared to healthy subjects were included.
Data extraction
Independent extraction of articles by 2 authors using predefined data fields, including study quality indicators.
Data synthesis
In total, the celiac patients had greater frequency of enamel defects (odds ratio = 5.69, 95%CI from 3.47 to 9.33, P < 0.00001, I 2 = 90%, 30 studies). Considering only the children, the odds ratio was 5.63 (95%CI from 3.95 to 8.01, P < 0.00001, I 2 = 65%, 24 studies), while in the adults the odds ratio was not significant (odds ratio = 2.16, 95%CI from 0.95 to 4.88, P = 0.06, I 2 = 40%, 3 studies). In total, the celiac patients had greater frequency of aphthous stomatitis (odds ratio = 3.79, 95%CI from 2.67 to 5.39, P < 0.00001, I 2 = 49%, 21 studies). Considering only the children, the odds ratio was 4.31 (95%CI from 3.03 to 6.13, P < 0.00001, I 2 = 29%, 13 studies), while in the adults the odds ratio was 47.90 (95%CI from 6.29 to 364.57, P = 0.0002, 1 study).
Conclusions
In children, celiac disease was associated with both enamel defects and aphthous stomatitis. The odds ratio estimates, however, should be interpreted with caution due to the high risk of bias showed by all the studies. In adults, the association between celiac disease and enamel defects or aphthous stomatitis was unclear because very few studies were performed on this population.
Clinical significance
The presence of enamel defects and/or aphthous stomatitis in a child affected by other typical or atypical symptoms of celiac disease represents an indication for further diagnostic exams for celiac disease.
1
Introduction
Celiac disease (CD) is a chronic small intestinal immune-mediated enteropathy precipitated by exposure to dietary gluten in genetically predisposed individuals . It affects about 1% of the population, two thirds of whom remain undetected and it is increasingly first diagnosed in adulthood .
Diagnosis of celiac disease is made through initial serologic testing and then confirmed by histopathologic examination of duodenal biopsies .
Celiac disease can present with many symptoms, including typical gastrointestinal symptoms such as diarrhea, steatorrhea, weight loss, bloating, flatulence, abdominal pain and also non-gastrointestinal abnormalities such as abnormal liver function tests, iron deficiency anemia, bone disease, skin disorders, and many other manifestations . Indeed, many individuals with celiac disease may have no symptoms at all. Generally, celiac disease is a benign disorder with a good prognosis in those who adhere to a gluten-free diet .
The oral cavity is one of the areas of celiac disease manifestation . Developmental dental defects of the dental enamel are especially frequently observed . Clinical picture of dental enamel structure defects in celiac patients may be diverse, and may contain hypoplasia, as well as hypomineralization . In recent years, many studies confirming a higher frequency of dental enamel defects in patients with celiac disease than in healthy subjects have been published .
Other complications observed in the oral cavity in patients with celiac disease are disturbances within mucosa . Aphthae are observed particularly often (RAS − recurrent aphthous stomatitis) .
Although many studies have been performed comparing the frequency of dental enamel defects and aphthae in celiac patient and healthy subjects, quantitative meta-analyses and systematic reviews are currently not being carried out.
The objective of this systematic review is to compare the frequency of enamel defects and aphthous stomatitis between celiac patients and healthy controls.
Following PICOS terms, the objective can be formulated as follows:
P (Population): Children (up to 18 years) and adults.
I (Interventions or Exposure): Patients with celiac disease.
C (Comparison or Control): Healthy individuals.
O (Outcomes): Presence/absence of enamel defects and presence/absence of aphthous stomatitis.
S (Study design): Controlled studies (cohort, cross-sectional, and case control studies).
This study is written in accordance with the PRISMA statement for reporting systematic reviews and meta-analysis of studies that evaluate health care interventions .
2
Materials and methods
2.1
Protocol and registration
This systematic review was registered on Figshare with the number 4665391.v1. The protocol can be accessed at: https://figshare.com/articles/Review_protocol_celiac_disease_and_enamel_defects_docx/4665391 .
2.2
Eligibility criteria
Only controlled studies on celiac patients (without other particular disorder like diabetes or Sjogren’s syndrome) compared to healthy subjects (without particular syndrome or disorder linked or not to celiac disease) were included. Children (up to 18 years) and adults were included.
No language or publication status restrictions were imposed.
Outcome variables were presence/absence of enamel defects and presence/absence of aphthous stomatitis at subject level in the two groups of individuals. Only clinical data were considered and histological data were not assessed.
2.3
Information source and study selection
Studies were identified by electronic database searches by two independent operators (MN, ET). No limits were applied for language. A systematic review of articles selected from MEDLINE (PubMed), EMBASE and Google Scholar was performed. Grey literature was searched on OpenGrey. The last search was run on 15 December 2016.
In addition, three reviewers (MN, ET, LF) hand-searched the pages of European Journal of Paediatric Dentistry, Pediatric Dentistry, Community Dentistry and Oral Epidemiology, Oral surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontics, Journal of Pediatric Gastroenterology and Nutrition, Gastroenterology, the American Journal of Gastroenterology, Journal of Gastroenterology, Journal of Dental Research, International Journal of Oral Science, together with abstracts printed in these journals from January 1990 to November 2016.
The reference lists of the previous reviews on this subject and of the selected articles were scanned to search for additional studies.
Eligibility assessment of the articles was performed in an independent manner by two reviewers (MN, ET). Any disagreement between reviewers was resolved involving a third operator (LF).
All full text articles selected were read by two reviewers (MN, ET) and articles that did not fulfil the inclusion criteria were excluded at this stage by consensus.
2.4
Search
The following search terms were used to search all databases: celiac disease, coeliac disease, dental enamel, aphthous.
The search strategy used in MEDLINE (Pubmed) was: (celiac[All Fields] AND (“dental enamel”[MeSH Terms] OR (“dental”[All Fields] AND “enamel”[All Fields]) OR “dental enamel”[All Fields] OR “enamel”[All Fields])) OR aphtous[All Fields] recurrent[All Fields] AND (“stomatitis, aphthous”[MeSH Terms] OR (“stomatitis”[All Fields] AND “aphthous”[All Fields]) OR “aphthous stomatitis”[All Fields] OR (“aphthous”[All Fields] AND “stomatitis”[All Fields])) AND celiac[All Fields] “coeliac disease”[All Fields] OR “celiac disease”[MeSH Terms] OR (“celiac”[All Fields] AND “disease”[All Fields]) OR “celiac disease”[All Fields] (“coeliac disease”[All Fields] OR “celiac disease”[MeSH Terms] OR (“celiac”[All Fields] AND “disease”[All Fields]) OR “celiac disease”[All Fields]) AND (“dental enamel”[MeSH Terms] OR (“dental”[All Fields] AND “enamel”[All Fields]) OR “dental enamel”[All Fields]) AND (“disease”[MeSH Terms] OR “disease”[All Fields]) .
2.5
Data collection process and data items
Two reviewers (MN, ET) extracted data from the included studies in a duplicate mode using a data extraction form. Disagreements were resolved by discussion between the two review authors. In cases of studies published more than once, information was used combining the papers.
Information was extracted from each included trial based on: title of the study, authors, year of publication, nationality of the first author, presence of blinded examiner, type of the study (cross-sectional, cohort, case-control), mean age of the subjects in the two groups, frequency of male and female in the two groups.
Variables were registered for each arm using frequency, and sample size. The variables were registered at patient level. When possible, enamel defects and aphthous stomatitis were registered separately for children (until 18 years old) and adults.
2.6
Risk of bias in individual studies
To ascertain the risk of bias of the selected controlled studies two reviewers (MN, ET) independently and blinded to each other used a simple ad hoc risk of bias tool. Disagreements were resolved by discussion between the two review authors.
The risk of bias tool consists of two items for which there is empirical evidence for their biasing influence on the estimates of an epidemiological trial. The two biases were the presence of a blinded examiner and the comparability of age and gender between the celiac patients and healthy individuals. If the two biases were satisfied the study was classified as a low risk of bias study. Alternately, if one of the bias was not satisfied the study was classified as a high risk of bias study.
2.7
Summary measures and planned method of analysis
The odds ratio of presence of dental enamel defects and the odds ratio of presence of aphthous stomatitis were used as a summary measure. The variables were registered at patient level.
Meta-analyses were performed using the Mantel-Haenszel method with random effect models. 95% confidence intervals for each outcome variable were calculated. The significance of any discrepancies in the estimates from different trials was assessed by means of the Cochran test for heterogeneity and the I 2 statistic, which describes the percentage total variation across studies that is due to heterogeneity rather than change.
Subgroup meta-analysis were performed for children (under 18 years old subject) and adults, and a third group consisting of the studies where the distinction between children and adults was not possible.
The statistical analyses were carried out using the RevMan software version 5.3 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) and the Software R version 3.3.1 (R foundation for statistical computing, Institute for statistic and mathematical, Vienna) by a single operator (MN).
2.8
Risk of bias across studies
Two funnel plots evaluating the asymmetry, which can result from the non-publication of small trials with negative results, assessed the possibility of publication bias. To test the asymmetry of the funnel plots (test for small effect) the Rücker’s test in the version of the arcsen of Thompson and Sharp was applied . If the Rücker’s test was significant (P < 0.05) the limit meta-analysis was applied to compensate for the bias .
2.9
Additional analyses
A pre-specified sensitivity analysis was performed including only low risk bias studies.
The influence analysis using the method leave-one-out was performed to verify the robustness of the results of the meta-analyses.
2
Materials and methods
2.1
Protocol and registration
This systematic review was registered on Figshare with the number 4665391.v1. The protocol can be accessed at: https://figshare.com/articles/Review_protocol_celiac_disease_and_enamel_defects_docx/4665391 .
2.2
Eligibility criteria
Only controlled studies on celiac patients (without other particular disorder like diabetes or Sjogren’s syndrome) compared to healthy subjects (without particular syndrome or disorder linked or not to celiac disease) were included. Children (up to 18 years) and adults were included.
No language or publication status restrictions were imposed.
Outcome variables were presence/absence of enamel defects and presence/absence of aphthous stomatitis at subject level in the two groups of individuals. Only clinical data were considered and histological data were not assessed.
2.3
Information source and study selection
Studies were identified by electronic database searches by two independent operators (MN, ET). No limits were applied for language. A systematic review of articles selected from MEDLINE (PubMed), EMBASE and Google Scholar was performed. Grey literature was searched on OpenGrey. The last search was run on 15 December 2016.
In addition, three reviewers (MN, ET, LF) hand-searched the pages of European Journal of Paediatric Dentistry, Pediatric Dentistry, Community Dentistry and Oral Epidemiology, Oral surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontics, Journal of Pediatric Gastroenterology and Nutrition, Gastroenterology, the American Journal of Gastroenterology, Journal of Gastroenterology, Journal of Dental Research, International Journal of Oral Science, together with abstracts printed in these journals from January 1990 to November 2016.
The reference lists of the previous reviews on this subject and of the selected articles were scanned to search for additional studies.
Eligibility assessment of the articles was performed in an independent manner by two reviewers (MN, ET). Any disagreement between reviewers was resolved involving a third operator (LF).
All full text articles selected were read by two reviewers (MN, ET) and articles that did not fulfil the inclusion criteria were excluded at this stage by consensus.
2.4
Search
The following search terms were used to search all databases: celiac disease, coeliac disease, dental enamel, aphthous.
The search strategy used in MEDLINE (Pubmed) was: (celiac[All Fields] AND (“dental enamel”[MeSH Terms] OR (“dental”[All Fields] AND “enamel”[All Fields]) OR “dental enamel”[All Fields] OR “enamel”[All Fields])) OR aphtous[All Fields] recurrent[All Fields] AND (“stomatitis, aphthous”[MeSH Terms] OR (“stomatitis”[All Fields] AND “aphthous”[All Fields]) OR “aphthous stomatitis”[All Fields] OR (“aphthous”[All Fields] AND “stomatitis”[All Fields])) AND celiac[All Fields] “coeliac disease”[All Fields] OR “celiac disease”[MeSH Terms] OR (“celiac”[All Fields] AND “disease”[All Fields]) OR “celiac disease”[All Fields] (“coeliac disease”[All Fields] OR “celiac disease”[MeSH Terms] OR (“celiac”[All Fields] AND “disease”[All Fields]) OR “celiac disease”[All Fields]) AND (“dental enamel”[MeSH Terms] OR (“dental”[All Fields] AND “enamel”[All Fields]) OR “dental enamel”[All Fields]) AND (“disease”[MeSH Terms] OR “disease”[All Fields]) .
2.5
Data collection process and data items
Two reviewers (MN, ET) extracted data from the included studies in a duplicate mode using a data extraction form. Disagreements were resolved by discussion between the two review authors. In cases of studies published more than once, information was used combining the papers.
Information was extracted from each included trial based on: title of the study, authors, year of publication, nationality of the first author, presence of blinded examiner, type of the study (cross-sectional, cohort, case-control), mean age of the subjects in the two groups, frequency of male and female in the two groups.
Variables were registered for each arm using frequency, and sample size. The variables were registered at patient level. When possible, enamel defects and aphthous stomatitis were registered separately for children (until 18 years old) and adults.
2.6
Risk of bias in individual studies
To ascertain the risk of bias of the selected controlled studies two reviewers (MN, ET) independently and blinded to each other used a simple ad hoc risk of bias tool. Disagreements were resolved by discussion between the two review authors.
The risk of bias tool consists of two items for which there is empirical evidence for their biasing influence on the estimates of an epidemiological trial. The two biases were the presence of a blinded examiner and the comparability of age and gender between the celiac patients and healthy individuals. If the two biases were satisfied the study was classified as a low risk of bias study. Alternately, if one of the bias was not satisfied the study was classified as a high risk of bias study.
2.7
Summary measures and planned method of analysis
The odds ratio of presence of dental enamel defects and the odds ratio of presence of aphthous stomatitis were used as a summary measure. The variables were registered at patient level.
Meta-analyses were performed using the Mantel-Haenszel method with random effect models. 95% confidence intervals for each outcome variable were calculated. The significance of any discrepancies in the estimates from different trials was assessed by means of the Cochran test for heterogeneity and the I 2 statistic, which describes the percentage total variation across studies that is due to heterogeneity rather than change.
Subgroup meta-analysis were performed for children (under 18 years old subject) and adults, and a third group consisting of the studies where the distinction between children and adults was not possible.
The statistical analyses were carried out using the RevMan software version 5.3 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) and the Software R version 3.3.1 (R foundation for statistical computing, Institute for statistic and mathematical, Vienna) by a single operator (MN).
2.8
Risk of bias across studies
Two funnel plots evaluating the asymmetry, which can result from the non-publication of small trials with negative results, assessed the possibility of publication bias. To test the asymmetry of the funnel plots (test for small effect) the Rücker’s test in the version of the arcsen of Thompson and Sharp was applied . If the Rücker’s test was significant (P < 0.05) the limit meta-analysis was applied to compensate for the bias .
2.9
Additional analyses
A pre-specified sensitivity analysis was performed including only low risk bias studies.
The influence analysis using the method leave-one-out was performed to verify the robustness of the results of the meta-analyses.
3
Results
3.1
Study selection
A total of 34 articles were identified for inclusion in the review (Andersson-Wenckert , Aine , Aguirre , Rea , Ventura , Latheenoja , Rasmusson , Sedghizadeh , Aydemir , Privolou , Bucci , Campisi , Procaccini , Wierink , Avsar , Campisi , Paez , Postek-Stefanska , Sezgin Bolgul , Malahias , Costacurta , Majorana , Ouda , El-Hodod , Ertekin , Acar , Shteyer , Bramanti , Alfar , Cantekin , De Carvalho , Herwis , Dane , Saraceno ).
The search on Medline provided a total of 89 citations, the search on the Embase provided 141 citations, the search on the Google Scholar provided 300 citations and the search on OpenGrey provided 2 citations ( Fig. 1 ).
An accurate review of the abstracts led to the selection of 46 articles from the electronic search and 2 additional studies from the hand-search. Two more studies were included from the bibliographical references of the articles identified.
Based on further analysis 16 of the 50 articles did not meet the inclusion criteria . Two studies were in vitro , 3 studies were not controlled , 3 studies had a control group with no healthy patients , 2 studies did not consider enamel defects or aphthae , 2 studies did not consider celiac patients , 3 studies presented incomplete data , 1 study was a double publication of the same data .
3.2
Study characteristics
All 34 trials selected for the review were controlled trials, 33 studies were published in English and one in Polish .
Ten trials were conducted in Italy , 7 in Turkey , 2 in Egypt , Finland , Spain , Sweden , USA , and one in Israel , Brazil , Holland , Greece , Jordan , Libya , and Poland .
The setting was public for 30 trials , private for 1 trial and unclear for 3 trials .
The design was cross-sectional for 32 studies , and case-control for 2 studies .
Twenty-two studies considered only children , 3 studies considered only adults , and 9 studies considered children and adults .
Fourteen studies dealt only enamel defects , 3 studies dealt only the aphthous stomatitis , and 17 studies dealt enamel defects and aphthous stomatitis .
3.3
Risk of bias within studies
Risk of bias within studies is shown for each article ( Table 1 ). All the studies presented at least one item at high risk of bias.
3.4
Results of individual studies and synthesis of results
3.4.1
Enamel defects
The 30 included studies involved 2636 celiac subjects and 10074 healthy subjects. In the individual studies the number of subjects varied between 38 to 7552 , and the odds ratios varied between 1.15 to 156.48 .
A statistically significant difference was obtained for the presence of enamel defects between celiac patients and healthy controls ( Fig. 2 ). In total, the celiac patients had greater frequency of enamel defects compared to healthy subjects. The odds ratio was 5.69 (95%CI from 3.47 to 9.33; P < 0.00001; I 2 = 90%, 30 studies). On 2636 celiac patients, 1083 (41%) had at least one tooth affected by enamel defects, while on 10074 healthy controls only 564 (6%) had at least one tooth affected by enamel defects.
Considering only the children, 1490 celiac subjects and 2318 healthy subjects were examined. The odds ratio was 5.63 (95%CI from 3.95 to 8.01, P < 0.00001; I 2 = 65%, 24 studies). On 1490 celiac patients, 686 (46%) had at least one tooth affected by enamel defects, while on 2318 healthy controls only 332 (14%) had at least one tooth affected by enamel defects.
Considering only the adults 174 celiac subjects and 337 healthy subjects were examined. The odds ratio was not significant (odds ratio = 2.16, 95%CI from 0.95 to 4.88, P = 0.06; I 2 = 40%; 3 studies). Of 174 celiac patients, 75 (43%) had at least one tooth affected by enamel defects, while of 337 healthy controls 134 (40%) had at least one tooth affected by enamel defects.
3.4.2
Aphthous stomatitis
The 21 included studies involved 1683 celiac subjects and 2162 healthy subjects. In the individual studies the number of subjects varied between 18 to 844 , and the odds ratios varied between 1.00 to 47.90 .
A statistically significant difference was obtained for the presence of aphthous stomatitis between celiac patients and healthy control ( Fig. 3 ). In total, the celiac patients had greater frequency of aphthous stomatitis (odds ratio = 3.79, 95%CI from 2.67 to 5.39, P < 0.00001, I 2 = 49%, 21 studies). On 1683 celiac patients, 420 (25%) had aphthous stomatitis, while on 2162 healthy controls only 232 (11%) had aphthous stomatitis.
Considering only the children, 1129 celiac subjects and 1408 healthy subjects were examined. The odds ratio was 4.31 (95%CI from 3.03 to 6.13, P < 0.00001; I 2 = 29%; 13 studies). On 1129 celiac patients, 686 (46%) had aphthous stomatitis, while on 1408 healthy controls only 118 (8%) had aphthous stomatitis.
Considering only the adults, 90 celiac subjects and 180 healthy subjects were examined. The odds ratio was 47.90 (95%CI from 6.29 to 364.57, P = 0.0002, 1 study). Of 90 celiac patients, 19 (21%) had aphthous stomatitis, while of 180 healthy controls only 1 (0.6%) had aphthous stomatitis.
3.5
Risk of bias across studies: funnel plot
3.5.1
Enamel defects
The funnel plot for the enamel defects is shown in Fig. 4 .
