What Is Known
- Some young genetically at-risk children for celiac disease may have transient antibody production.
- Antitransglutaminase antibody levels are significantly lower in children with potential celiac disease and celiac autoimmunity than in those with an intestinal lesion.
What Is New
- Some young, scarcely symptomatic, genetically at-risk children for celiac disease with high levels of antitransglutaminase and endomysial antibodies have a normal intestinal biopsy.
- The prevalence studies of celiac disease, especially those carried out in young children, should be based on intestinal damage in the intestinal biopsy so as to not overestimate results.
Celiac disease (CD) is an immune-mediated systemic disorder elicited by gluten and related prolamines in genetically susceptible individuals and characterized by the presence of a variable combination of gluten-dependent clinical manifestations, CD-specific antibodies, HLA-DQ2 or -DQ8 haplotypes, and enteropathy (1). A gluten-free diet leads to a prompt clinical improvement and the histological recovery of the damaged mucosa.
The development of highly sensitive and specific serologic tests has allowed us to estimate the prevalence of the disease in 1% of the general population (2). Moreover, screening programs within populations have contributed to the recognition of atypical and silent forms of the disease that can remain underdiagnosed. Untreated CD is linked to significant morbidity later in life such as osteoporosis (3), T-cell non-Hodgkin lymphoma (4), and possibly other autoimmune diseases (5). Most studies have estimated the prevalence of the disease in the general population. Studies assessing the prevalence of CD in populations selected by genetic susceptibility, however, are scarce (6,7).
Genes encoding HLA-DQ2 or -DQ8 are found in most patients with CD (8), and the absence of these haplotypes virtually rules out the disease. Approximately 30% of the white population also, however, carries HLA-DQ2 (9) and the majority do not develop CD. It is unclear why CD develops in only a minority of genetically susceptible individuals, even though nearly all individuals in Western populations are exposed to gluten. This together with the different and changing patterns of the disease indicates that other factors, in addition to gluten, may play a role. Several epidemiological studies have shown an association between CD and environmental factors such as viral or bacterial infection (10), the pattern of breast-feeding (11), the timing of gluten introduction into the diet (12), and mode of delivery (13), although these findings are still controversial (14,15). Moreover, the studies controlled for the HLA genotype did not provide evidence on the protective role of breast-feeding in the development of CD (6,7,12,16,17). Follow-up of at-risk cohorts allowed us the opportunity to study different epidemiological factors in patients and referents with the same genetic background.
The aim of the present study was to assess the prevalence and clinical presentation of CD in a cohort of children carrying the HLA-DQ2 haplotype and evaluate environmental factors other than gluten that could influence the risk of developing CD in this population.
Between July 2004 and July 2005, the parents of all healthy full-term newborns in our hospital were invited to participate in the study. Follow-up of all of the children at risk for CD (HLA-DQ2 positive) was performed. Our hospital has a well-defined catchment area with a pediatric gastroenterology unit and no pediatric-specialized private facilities.
In order to create a cohort of at-risk children for CD who could be studied prospectively with time, we tested the HLA-DQ2 haplotype in a blood sample of the umbilical cord. In this cohort, 2 groups of children were studied. The screening group, asymptomatic children called in for a serological test when they were between 2 and 3 years of age; and the symptomatic group, children with symptoms of CD, before the screening study started. In the screening group, a point-of-contact test (POC test) was performed, and during the visit, a registration form was filled out containing age (months), sex, and mode of delivery (vaginal or cesarean). Data about breast-feeding (>15 days), breast-feeding duration (months), and age of gluten introduction (months) were retrospectively collected. In children with a positive POC test, a blood sample was collected to confirm these results by serum anti-transglutaminase 2 (anti-TG2) and endomysial antibodies (EMA). Children with serum-positive antibodies were referred to our pediatric gastroenterology unit for a complete clinical evaluation and were retested about 3 months later. Children with high antibody levels, anti-TG2 ≥80 (>10-fold the upper normal value) and/or EMA ≥1:80 underwent an intestinal biopsy. Children with lower antibody values were followed up with the serological test every 6 months. The symptomatic group consisted of children attended by their primary-care pediatricians who suspected the disease and requested CD serology that proved the positivity of the antibodies. These children were then referred to our pediatric gastroenterology unit for confirmatory endoscopy. When this study was performed, the usual way to screen for symptomatic CD in our hospital was to determine only EMA. Therefore, we do not have any data on TG2 antibodies in this group of children (Fig. 1).
The HLA-DQ2 haplotype was tested in a blood sample of the umbilical cord by polimerase chain reaction and sequence-specific oligonucleotide probing (PCR-SSOP) as earlier described (18). The result was considered positive if both alleles, HLA-DQA1*05 and -DQB1*02, were positive. This technique enables us to ascertain that both alleles are present but fails to provide information regarding the haplotype in which they are included as well as the number of copies of each allele. The results are positive in patients who have the DQ2 heterodimer in cis position with a genotype DQA1*0501-DQB1*0201/DDQ*XX-DQB*X′X′ (DQ2.5/DQX′), in trans position with a genotype DQA1*0505-DQB1*0301/DQA1*0201-DQB1*0202 (DQ7/DQ2.2), and homozygotes for the DQ2.5 (DQ2.5/DQ2.5).
Anti-TG2 antibody detection (immunoglobulin [Ig]A/IgG/IgM) was performed in a whole blood sample taken from a finger prick at the point of contact using a rapid test kit (POC test) following the manufacturer's instructions (Operon-Tec-Laim S.A, Zaragoza, Spain). These results were confirmed by serum anti-TG2 and EMA determination. Anti-TG2 levels were measured by an enzyme-linked immunosorbent assay method based on human recombinant antigen, using a commercial kit (Celikey. ImmunoCap250 Phadia, Uppsala, Sweden); values ≥8 were considered to be positive. EMA values were determined by a commercial indirect immunofluorescence assay with monkey esophagus as the substrate (BioSystems, Atom, Barcelona, Spain); a dilution ≥1:5 was considered to be positive. Positive sera were further diluted 1:10, 1:20, 1:40, 1:80, and 1:160. The following laboratory values were also measured by standard laboratory methods: hemoglobin level (normal values ≥11.5 g/dL), ferritin level (normal values ≥12 ng/mL), liver enzymes levels (aspartate aminotransferase [AST]: 20–39, alanine aminotransferase [ALT]: 0–40, gamma-glutamyl transpeptidase [GGT]: 11–49 reference values), and total serum IgA. IgA deficiency was defined at serum levels <0.2 g/L.
Celiac Disease Diagnosis
Symptoms and signs of CD included chronic or intermittent diarrhea, weight loss, abdominal distension, failure to thrive, chronic constipation, vomiting, irritability, iron-deficiency anemia, and increased levels of liver enzymes. An intestinal biopsy was performed by upper gastrointestinal endoscopy with 1 to 2 specimens obtained from the bulb and 3 to 4 from the descending part of the duodenum. The small intestinal damage was graded according to the Marsh-Oberhuber classification (19,20). Children with grade-II (intraepithelial lymphocytosis and enlarged crypts), -IIIa (partial villous atrophy), -IIIb (subtotal villous atrophy), and -IIIc (total villous atrophy) were diagnosed with CD.
Study of Risk Factors
We performed a case-referent study to assess the role of mode of delivery, breast-feeding, breast-feeding duration, and age of gluten introduction in the development of CD in this population. The cases were the children with CD confirmed by intestinal biopsy, and referents were all children of the HLA-DQ2–positive cohort but with negative CD antibodies.
Proportions were expressed as percentages and 95% confidence interval (CI). Continuous variables, presented as mean ± standard deviation, were compared between groups by using the t test. Differences in the frequency of independent variables between patients and referents were assessed by χ2 with continuity correction. The odds ratio (OR) and 95% CI were calculated to estimate the strength of the association. All P values were two-sided and values <0.05 were considered statistically significant. SPSS version 14.0 (SPSS Inc. Chicago, IL) was used for the statistical analyses.
The study protocol was approved by the Clinical Research Ethics Committee of Severo Ochoa Hospital (Leganés, Madrid). Written informed consent, which included information about future serological testing, was obtained from the parents of each infant upon entering the study.
HLA-typing and Serological Study
Between July 2004 and July 2005, 1716 children were born in our hospital. Parents of 1291 (75.2%) (620 girls) agreed to participate in this prospective follow-up. We were unable to contact parents of 13.1% of the eligible children, usually because of the early hospital discharge, and 11.7% of parents who were asked to participate refused the screening, mainly because of the linguistic problems. Of the 1291 children who joined the study, 362 (28%, 95% CI 25–30) (44.5% girls) carried the HLA-DQ2 haplotype. Of these 362 children, 262 (72.3%) participated in the serological screening; 255 were recruited between the ages 2 and 3 (screening group), and 7 were referred from primary-care pediatricians to our pediatric gastroenterology unit with symptoms of CD and positive antibodies before the serological screening started (symptomatic group). The median age of the screening group and the symptomatic group was 29.3 months (25–38 months) and 20.7 months (16–24 months), respectively.
In the screening group, the POC test was positive in 20 of 255 children. These results were confirmed in 19 by serum determination of IgA–anti-TG2 and IgA-EMA antibodies. All of the 19 children tested positive to both antibodies in the retesting 5.2 % (95% CI 2.8–7.7) (13 girls) and were suspected of having CD. Taking into account 7 additional children who tested positive in the symptomatic group, 26 out of 362 of the whole HLA-DQ2–positive cohort, 7.1% (95% CI 4.4–9.9) (18 girls) had antibodies related to CD.
Twenty children, 14 in the screening group and 6 in the symptomatic group, fulfilled the requirements for intestinal biopsy. CD was confirmed in 15 children (12 girls), 9 from the screening group and 6 from the symptomatic group, and 5 had a normal biopsy. CD was diagnosed in 15 of the 362 children with HLA-DQ2 positive, 4.1% (95% CI 1.9–6.3) or 1 of 24 children. If we allude to the general population, which includes the newborns enrolled, the prevalence of the disease was 1.1% (95% CI 0.54–1.78) or 1 of 86 children. The results are summarized in Fig. 1.
Symptoms and Signs
Clinical characteristics and antibody levels of the screening group and the symptomatic group are presented in Table 1. In the total population, 60% had gastrointestinal symptoms, 7% poor weight gain, and 33% were asymptomatic. In the screening group, 56% were asymptomatic, 33% showed abdominal distension, and 11% gastrointestinal symptoms (intermittent diarrhea). In the symptomatic group, 83% showed gastrointestinal symptoms (classical forms) and 17% weight loss without any other symptom. None of the children had IgA deficiency. All of them had normal transaminase levels. Iron deficiency was found in 7 of 15, and 5 (33.3%) had iron deficiency anemia, 2 in the screening group, and 3 in the symptomatic group.
The serological follow-up of the 5 children with positive antibodies and normal intestinal biopsy (potential CD), and the 6 children with low antibody levels and without intestinal biopsy, or, in keeping with present definitions, children with CD autoimmunity (21) is shown in Fig. 2. We observed a progressive loss of the antibodies with time until all the children had negative results. A total of 11 children out of 26 (42.3%) with positive antibodies became negative and are still negative on a normal diet after 5 to 7 years of follow-up. Ten of these children belonged to the screening group. The average time during which antibodies were lost was 14.5 months (4–22 months), 19.2 months in children with potential CD, and 10.6 months in children with CD autoimmunity.
We compared the 15 children with CD with the 220 children of the same cohort but with negative antibodies. In the celiac group, we observed a higher percentage of girls, caesarean delivery, family history of CD, and breast-feeding, but only in the case of girls was there a significant difference between the 2 groups (OR 5.7, 95% CI 1.49–19,9). On the contrary, gluten introduction during breast-feeding was significantly more frequent in the referent group (OR 0.11, 95% CI 0.01–0.8). The duration of breast-feeding was longer in referents (6.5 ± 6.7 months) than in children with celiac disease (4.8 ± 6 months), but this difference was not significant (P = 0.81).
In our cohort of children with HLA-DQ2 positive between 2 and 3 years of age from the general population, the prevalence of CD-positive antibodies was 7% and the CD confirmed by intestinal biopsy was 4%. In the whole cohort, considering the children detected by screening and those referred to us by their primary-care pediatricians with suspected CD, more than half of the children had gastrointestinal symptoms. In children in whom CD could not be confirmed, we observed a spontaneous negativization of both anti-TG2 and EMA antibodies, which persisted after 5 to 7 years of follow-up.
We chose, as the first step in the serological study, a POC test which was quick, simple to perform, and easy to interpret. Anti-TG2 antibody detection by POC test proved to be a valid tool in the diagnosis of CD with >95% sensitivity and specificity and in total concordance with the intestinal lesion when the test was performed on populations with high prevalence of CD (22). When the prevalence of CD is lower, however, the positive predictive value drops to approximately 70% and the negative predictive value rises to approximately 100% (23). It has been shown that the POC test seemed to be as accurate as laboratory testing and had high positive and negative predictive values and specificity in a primary-care CD screening study on 6-year-old children. Sensitivity, however, was low when faint lines were interpreted as negative results (24). In our study, the only test with a faint line turned out to be negative by serum determination of anti-TG2 and EMA antibodies.
In a study involving a large number of children carrying the HLA-DQ2 and/or -DQ8 haplotypes, 12% had anti-TG2 antibodies by the age of 5 years, which increased to 26% in children homozygous for the HLA-DQ2 haplotype (25). In our study, 7% of the children with HLA-DQ2 positive had anti-TG2 and EMA-positive antibodies. But without any data about the number of copies of every allele, we were only able to ascertain the prevalence of autoantibodies in the whole cohort. Twenty children fulfilled the requirements for intestinal biopsy, and CD was eventually confirmed in 15 children, 9 from the screening group and 6 from the symptomatic group, which means a minimum prevalence of 4.1% of the children with HLA-DQ2 positive because we were not able to contact with all of the parents of the children with DQ2 positive. If we allude to the general population, which includes all the newborns enrolled, the prevalence of the disease was 1.1%. Because of methodological differences, an accurate comparison with studies performed in at-risk children is difficult. Bjorck et al (26) reported a prevalence of 3.5% in a cohort of young children carrying the HLA-DQ2 and -DQ8 haplotypes. This figure increased to nearly 5% when children with a previous diagnosis of CD were included. In this study, however, the number of celiac patients could be overestimated because patients with a Marsh 1 lesion were diagnosed with CD. Our figure is similar to the prevalence of 4.5% in schoolchildren carrying the HLA-DQ2 haplotype (27). In 2 recent studies in children positive for HLA-DQ2 or -DQ8 and with at least 1 first-degree relative with CD, the cumulative incidence of CD was 5.2% among children at 3 years of age in the Prevent CD study (7), and approximately 6% among children at 2 years of age in the Celiprev study (6).
More than 50% of the celiac patients showed gastrointestinal symptoms with or without iron-deficiency anaemia, 7% poor weight gain without any other gastrointestinal symptoms, and 33% were asymptomatic. If we consider the 9 patients diagnosed by screening, 5 were asymptomatic, and in only 1 case did parents report CD-related symptoms (intermittent diarrhea). After a physical examination and blood tests, 3 showed abdominal distension with iron deficiency with or without anemia. Taking the CD clinical spectrum as an “iceberg,” we observed that 60% of the patients were diagnosed during the screening and, therefore, were below water level and the 33% and 7% of classic and nonclassic forms, respectively, represented the visible part of the “iceberg.” These figures are similar to those found in a screening program on primary schoolchildren in Italy (28).
During the follow-up, the 5 children with potential CD and the 6 children with celiac autoimmunity lost their antibodies without having made any dietary changes. This accounts for the high percentage (42.3%) of children who became antibody-negative in the total cohort. All but 1 of these children came from the screening group. We have no reason to believe that this negativization is a consequence of parents putting their children on low gluten diet, especially considering the lack of symptoms, the added costs and our insistence on maintaining a normal diet. Previous studies have shown that some young genetically at-risk children may have transient antibody production (3,29,30) but the antibody levels were only moderately elevated as had occurred in the 6 children with transitory CD autoimmunity in our study. Not all the children with high antibody levels, however, showed small intestinal damage. All but 1 with potential CD showed anti-TG2 levels 10 times the upper normal limit. Moreover, the EMA titres of 3 of these 5 children were of 1:80 or more although it had been previously demonstrated that these EMA titres together with high anti-TG2 levels tended to persist and predict a more severe histological lesion (31). In a recent mass screening in a population of 12-year-olds, all the patients with anti-TG2 ≥10 times the upper limit of normal had gluten enteropathy (32). In a previous screening study done by our group on a population of schoolchildren between 10 and 12 years of age, only 1 out of 15 children with positive EMA antibodies had a normal biopsy (33). It seems that, in screening studies, high antibody levels in schoolchildren correlated better with intestinal damage than in younger children. Therefore, in young scarcely symptomatic, genetically at-risk children, in whom there seems to be a more marked transitory autoimmunity, time should be given to antibody evolution before carrying out biopsies to avoid negative results. In accordance with the published data, extending the retesting period from 3 to 6 months could have saved some normal biopsies in our group of children with potential CD (34). The overestimation of the prevalence of CD when anti-TG2 antibodies were used as the only diagnostic test was demonstrated in a systematic review. This was linked to the lower specificity of these antibodies compared with EMA (35). In our study, we tested both antibodies to take advantage of the better specificity of EMA but these antibodies also turned out to be positive in all the children with potential CD. Studies on the prevalence of CD, especially those carried out in young children, even with high levels of both antibodies, should be based on mucosa damage in the intestinal biopsy to not to overestimate the results. In our study, after 5 to 7 years of follow-up, all the children that became antibody-negative are still negative on a normal diet. We continue to monitor these children but on a less frequent bases.
In order to assess the role of risk factors, other than gluten, in the development of the disease, the follow-up of our cohort enabled us to compare biopsy-confirmed celiac patients with children with negative antibodies. In our study, we observed a higher percentage of girls, cesarean delivery, family history of CD, and breast-feeding but only in the case of girls was there a significant difference between groups. Primary-care pediatricians in our hospital catchment area follow similar protocols when promoting breast-feeding and introducing complementary feeding. At the time of the study, gluten introduction was between 6 and 7 months of age. The duration of breast-feeding, although longer in referents, was not significantly different from celiac patients, probably because of the small sample size. In our series, those children who were given gluten during breast-feeding showed less risk of developing CD. This fact, found in retrospective studies, carried out in celiac patients (36), has not been observed in studies controlled for the HLA haplotype (12,16). Moreover, 3 recent studies involving a significant number of children at risk of CD have shown that infant-feeding practices, breast-feeding duration, and time of gluten introduction, have no effect on the risk of developing CD. These studies confirmed that the HLA represents the main risk factor with a gene-dose effect on the development of the disease (6,7,17). In our study, data related to risk factors were retrospectively collected with the potential for parental recall bias. The low number of cases and the wide CI of the OR make our findings regarding the risk factors inconclusive.
Other limitations of this study should be noted. We were not able to determine the HLA-DQ8 owing to a lack of personnel and economic resources. We consider that the inclusion of this haplotype, however, would not have substantially modified the final results. In a national registry of CD in our country, only 2.3% of the celiac patients carried the HLA-DQ8 haplotype (37). After at least 2 years without any contact with the families, we were unable to locate 28% of HLA-DQ2 positive. We considered the total number of children with HLA-DQ2 positive when calculating the prevalence of CD in this cohort. Therefore, the figures obtained are an estimate of the minimum number of celiac patients in our study population. We did not perform the extended HLA-DQ2 haplotype and were unable to identify children with the highest genetic risk.
In conclusion, the prevalence of CD in this cohort of children with HLA-DQ2 positive in their first 3 years was 4.1%. In the whole cohort, more than half of the patients had gastrointestinal symptoms with or without iron deficiency anemia, 7% poor weight gain, and 33% of them showed no signs and symptoms related to CD. As a high proportion of children showed a spontaneous disappearance of both antibodies, prevalence studies of CD in young children should be based on intestinal damage so as to not overestimate results.
1. Husby S, Koletzko S, Korponay-Szabó IR, et al. European Society for Pediatric Gastroenterology, Hepatology, and nutrition guidelines for the diagnosis of coeliac disease. J Pediatr Gastroenterol Nutr
2. Dubé C, Rostom A, Sy R, et al. The prevalence of celiac disease in average-risk and at risk Western European populations: a systematic review. Gastroenterology
2005; 128 (4 suppl 1):S57–S67.
3. Agardh D, Bjorck S, Agardh CD, et al. Coeliac disease specific tissue transglutaminase antibodies are associated with osteoporosis and related fractures in middle-age women. Scand J Gastroenterol
4. Tio M, Cox MR, Eslick GD. Meta-analysis: coeliac disease and the risk of all-cause mortality, any malignancy and lymphoid malignancy. Aliment Pharmacol Ther
5. Ventura A, Maggazu G, Greco L. Duration of exposure to gluten and risk for autoimmune disorders in patients with celiac disease. Gastroenterology
6. Lionetti E, Castellaneta S, Francavilla R, et al. Introduction of gluten, HLA status, and the risk of celiac disease in children. N Engl J Med
7. Vriezinga SL, Auricchio R, Bravi E, et al. Randomized feeding intervention in infants at high risk for celiac disease. N Engl J Med
8. Sollid L, Thorsby E. HLA susceptibility genes in celiac disease: genetic mapping and role in pathogenesis. Gastroenterology
9. Louka A, Moodie S, Karell K, et al. A collaborative European search for non-DQA1-05-DQB1-02 celiac disease loci on HLA-DR3 haplotypes: analysis of transmission from homozygous parents. Hum Immunol
10. Plot L, Amital H. Infectious associations of celiac disease. Autoimmun Rev
11. Akonberg AK, Ramanan AV, Buchan I, et al. Effect of breast feeding on risk of coeliac disease: a systematic review and meta-analysis of observational studies. Arch Dis Child
12. Norris JM, Barriga K, Hoffenberg EJ, et al. Risk of celiac disease autoimmunity and timing of gluten introduction in the diet in infants at increased risk of disease. JAMA
13. Decker E, Engelmann G, Findeisen A, et al. Cesarean delivery is associated with celiac disease but not inflammatory bowel disease in children. Pediatrics
14. Welander A, Tjernberg AR, Montgomery SM, et al. Infectious disease and risk of later celiac disease in childhood. Pediatrics
15. Størdal K, White RA, Eggesbø M. Early feeding and risk of celiac disease in a prospective birth cohort. Pediatrics
16. Ascher H, Krantz I, Rydberg L, et al. Influence of infant feeding and gluten intake on celiac disease. Arch Dis Child
17. Aronsson CA, Lee HS, Liu E, et al. Age at gluten introduction and risk of celiac disease. Pediatrics
18. Olerup O, Aldener A, Fogdell A. HLA-DQB1 and -DQA1 typing by PCR amplification with sequence-specific primers (PCR-SSP) in 2 hours. Tissue Antigens
19. Marsh MN. Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity (‘celiac sprue’). Gastroenterology
20. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of coeliac disease: time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol
21. Ludvigsson JF, Leffler DA, Bai JC, et al. The Oslo definitions for coeliac disease and related terms. Gut
22. Raivio T, Kaukinen K, Nemes E, et al. Self transglutaminase-based rapid celiac disease antibody detection by lateral flow method. Alim Pharmacol Ther
23. Giersiepen K, Lelgemann M, Stuhidreher N, et al. Accuracy of diagnostic antibody for coeliac disease in children: summary of an evidence report. J Pediatr Gastroenterol Nutr
24. Korponay-Szabo IR, Szabados K, Pusztai J, et al. Population screening for celiac disease in primary care by district nurses using a rapid antibody test: diagnostic accuracy and feasibility study. BMJ
25. Liu E, Lee HS, Aronsson CA, et al. Risk of pediatric celiac disease according to HLA haplotype and country. N Engl J Med
26. Bjöck S, Brundin S, Lörinc E, et al. Screening detects a high proportion of celiac disease in young HLA-genotyped children. J Pediatr Gastroenterol Nutr
27. Mäki M, Mustalahti K, Kokkonen J, et al. Prevalence of celiac disease among children in Finland. N Engl J Med
28. Nenna R, Tiberti C, Petrarca L, et al. The celiac iceberg: characterization of the disease in primary schoolchildren. J Pediatr Gastroenterol Nutr
29. Simell S, Pupila A, Hoppu S, et al. Natural history of transglutaminase autoantibodies and mucosal changes in children carrying HLA-conferred celiac disease susceptibility. Scand J Gastroenterol
30. Lionetti E, Castellaneta S, Pulvirenti A, et al. Prevalence and natural history of potential celiac disease in at-family-risk infants prospectively investigated from birth. J Pediatr
31. Agardh D, Carlsson A, Lynch K, et al. Usingradioligand-binding assays to measure tissue transglutaminase autoantibodies in young children. Acta Paediatr
32. Sandström O, Rosén A, Lagerqvist C, et al. Transglutaminase IgA antibodies in a Celiac Disease mass screening and the role of HLA-DQ2 Genotyping and endomisial antibodies in sequential testing. J Pediatr Gastroenterol Nutr
33. Cilleruelo ML, Román E, Jiménez J, et al. Silent celiac disease: exploring the iceberg in school-aged population. An Esp Pediatr
34. HogenEsch CE, Csizmadia GDS, VanHoogstraten IMW, et al. Childhood coeliac disease: towards an improved serological mass screening strategy. Aliment Pharmacol Ther
35. Biagi F, Klersy C, Balduzzi D, et al. Are we not over-estimating the prevalence of coeliac disease in the general population? Ann Med
36. Ivarsson A, Hernell O, Stenlund H, et al. Breast-feeding protects against celiac disease. Am J Clin Nutr
37. Cilleruelo Pascual ML, Román Riechmann E, Sánchez Valverde F, et al. Spanish national registry of Celiac Disease: incidence and clinical presentation. J Pediatr Gastroenterol Nutr