What Is Known
- Despite a greater awareness and increased availability of serological tests, many celiac disease patients are still undiagnosed.
- The prevalence of celiac disease in Denmark has previously been reported to be very low and has been estimated to 0.08% in Danish children compared with a worldwide prevalence of 1%.
What Is New
- The prevalence of celiac disease in Danish children and adolescents is much higher than previously reported.
- In this population-based cohort, we estimate the prevalence of celiac disease of 2.6% where more than half of them were found by screening.
Celiac disease (CeD) is an autoimmune disease that primarily affects the small intestine, and it is unique in that the triggering antigen “gluten” is known (1). The clinical presentation of CeD varies from severe pediatric cases with failure to thrive, malabsorption, and diarrhea to an asymptomatic presentation (2).
Despite the increased awareness and availability of serological tests, many individuals with CeD remain undiagnosed. A study from 2010 found a 1.6% prevalence of undiagnosed CeD in a random sample of the general population (3). Undetected and hence untreated CeD may have a higher risk of morbidities, such as cancer and osteoporosis (4), and even mortality in very long follow-up (5). Only a few studies have, however, examined the long-term consequences of undiagnosed CeD.
The prevalence of diagnosed CeD in Denmark has traditionally been regarded as very low (6). In 2010, the prevalence of diagnosed CeD in children was, however, estimated to be 84 per 100,000 children ages 0 to 17 years, a considerable increase from 43 per 100,000 children in 1996 (7). This was supported by a Danish screening study, which found a 0.5% prevalence of CeD in adults (8), which is lower than other western countries and much lower than neighboring Sweden (3% in children and 2% in adults) (9,10). A screening study of Danish and Swedish blood donors from 1996, however, revealed almost the same prevalence of celiac seropositivity (0.25% and 0.27%, respectively), which suggests that the prevalence of CeD in Denmark is similar to that of Sweden, and is thus underdiagnosed in Denmark (11).
The aim of this present study is to provide an estimated prevalence of screen-detected CeD in adolescents, an age group, which has never been specifically studied.
The Study Population
The Danish National Birth Cohort (DNBC) consists of approximately 96,000 children who are followed from intrauterine life onwards, with the most recent follow-up at the age of 18 years (12). The mothers of these children were recruited from 1996 to 2002 from the general population with a participation rate of approximately 30%.
The Glutenfunen cohort was recruited from an unselected subsample of the Danish National Birth Cohort, defined as those living on the Island of Funen, Denmark. Funen is the second largest island in Denmark and consists of both rural and urban environments with a sociodemographic profile that is representative of Denmark as a whole (13). Funen has a total population of about 500,000 (40,000 in the age range of 15–21 years), corresponding to approximately 9% of Denmark's population. They constituted 7431 participants in the age range of 15 to 21 years.
Nonparticipants were defined as those invited from the DNBC subsample who either declined or failed to respond to the invitation.
Eligible participants in the Glutenfunen cohort received an invitation to a clinical visit via their digital mailbox, which is established for all Danish residents, followed by a written letter. For those under 18 years of age, the invitation was sent to the mother. If the individual agreed to participate, a pediatrician (C.C.) called them or their mother (if age < 18 years) to provide oral information and obtain oral consent. Around 24 hours later, the participants received a link in their digital mailbox to provide written consent. Only participants who provided both oral and written consent were included. A short questionnaire and fecal sample collection kit were then sent to the participant.
The recruitment process took place from May 31, 2017 to July 2020, interrupted by the COVID-19 crisis from March 9 to June 1, 2020.
The clinical visit took place at Hans Christian Andersen Children's Hospital, Odense University Hospital, or at the affiliated Svendborg Hospital. During the clinical visit, anthropometric measurements were taken, and blood samples and a fecal sample were collected. Height, weight, and BMI were reported as z-scores to allow for age and sex differences using the Danish references for growth (14). Age was defined as the age on the day participants received oral information. Blood pressure (BP) was measured after 5 minutes of total rest using the Omron-HBP-1100 automatic BP apparatus. The lowest BP was reported and expressed in z-scores (15).
The blood samples were tested for the following CeD antibodies: transglutaminase-2 IgA (TG2-IgA) and deamidated gliadin peptide IgG using assay from INOVA Diagnostic, Inc., San Diego, CA and total IgA using Optilite kit from The Binding Site Group Limited, UK. If total IgA was less than 0.5 g/L, transglutaminase-2 IgG and endomysial IgG antibodies were measured also using assays from INOVA. Furthermore, HbA1C and a lipid profile were measured.
Positive Transglutaminase-2 IgA
Participants with TG2-IgA higher than the reference range were invited to a clinical visit, where a gastroscopy with biopsies under general anesthesia was offered, in accordance with European Society for Paediatric Gastroenterology Hepatology and Nutrition (ESPGHAN) guidelines at the time (1). If the participant declined gastroscopy, retesting of a new blood sample, including IgA endomysial antibodies (EMA-IgA), was offered.
CeD was diagnosed if the biopsy showed changes corresponding to Marsh II or III of the Marsh classification (16). Potential CeD was defined as positive Tg2 antibodies without a positive histological pattern in the duodenal biopsy. Known CeD was recorded based on the questionnaire, and was then verified in the electronic medical records by author C.C. For nonparticipants, diagnosed CeD was based on the Danish National Patient Registry, which contains information about all inpatient and outpatient visits to Danish hospitals. The diagnosis of the International National Classification of diseases 10th revision (ICD-10) for CeD (K90.0) was used (17). Diagnosed CeD was defined as at least 2 registrations in the Danish National Patient Registry to avoid the inclusion of false registrations. Data extraction was done on December 31, 2013.
The prevalence in the Glutenfunen cohort was defined as the number of CeD cases diagnosed before and during this study divided by the number of participants in the Glutenfunen cohort. The minimum prevalence of CeD was defined as the number of CeD cases diagnosed before and during this study out of all invited participants, assuming that no other undiagnosed cases existed among the nonparticipants.
Study data were collected and managed using REDCap (Research Electronic Data Capture) electronic data capture tools hosted at University of Southern Denmark (18).
The project was approved by the Regional Committee for Biomedical Research in the Region of Southern Denmark (project no S-20160061) and the Danish Data Protection Agency (2008-58-0018). A specific statement of consent for the biobank was obtained from each participant and/or by a parent. All authors had access to the study data and reviewed and approved the final manuscript.
We included 1266 participants out of 7431 eligible participants (17%) in the Glutenfunen cohort. Of the included participants, 815 (64%) were under age 18 and 451 (36%) were 18 years or older. The participation rate was higher in the group <18 years old who were invited through the mothers. The distribution of gender was 811 (64%) girls and 455 (36%) boys compared with 2868 girls (47%) and 3297 (53%) boys in the nonparticipating group. There was no difference in age between the Glutenfunen cohort participants compared with nonparticipants. Age for nonparticipants was defined as the age of the participant on August 1, 2017. A total of 32% of the participants lived in Odense, the main city on Funen, compared with 30% in the nonparticipant group. Participants with known CeD were more eager than those without known CeD to participate in the Glutenfunen cohort (see Fig. 1 and Table 1). Blood samples were provided by 1173 (93%) participants.
TABLE 1 -
Basic characteristics of the Glutenfunen Cohort compared with the Danish National Birth Cohort on Funen
||Glutenfunen cohort, n = 1266
||Non-participants from the DNBC subsample on Funen, n = 6165
| Females, n (%)
| Males, n (%)
|Age, years‡ (25%, median, 75%)
||16.6 17.5 18.4
||16.6 17.6 18.5
| <18, n (%)
| ≥18, n (%)
| City, n (%)
| Rural area (%)
| Known Ced||
Ced = celiac disease; DNBC = Danish National Birth Cohort.
∗Values are mean for normally distributed variables, median and percentiles for nonnormally distributed variables and n (valid percent) for categorical variables. If the values differed significant P < 0.05 there are marked with an asterisk (∗).
†based on the Danish national personal identification number (CPR).
‡The age was calculated as the age for the participants August 1st 2017.
§On the basis of data from Statistics Denmark.
||On the basis of the questionnaire to the participants and from Danish National Patient Registry.
Prior Celiac Disease
The prevalence of diagnosed CeD was 0.4%. On the basis of the questionnaire, 14 participants (1.1%) had known CeD before entering the Glutenfunen cohort. From the Danish National Patient Register, we found 14 cases with known CeD among the nonparticipants (0.2%).
Screen-detected Celiac Disease
Thirty-three participants were positive for TG2-IgA. Nineteen participants had duodenal biopsies, which showed changes compatible with CeD. Eleven of these had TG2-IgA above or at 10 times the reference range. All participants were evaluated by clinicians (C.C. or S.H.) before endoscopy and among the 19 participants with a biopsy compatible with CeD, 1 had moderate and 5 had mild gastrointestinal symptoms. The rest were asymptomatic. Ten participants had a normal biopsy and 4 participants refused endoscopy but provided a new blood sample, whilst still on a gluten-containing diet, and all 4 were negative for EMA-IgA and had low TG2-IGA titers. The lowest value of TG2-IgA and a biopsy compatible with CeD was 23 U (the reference range was <20 U).
Thirteen participants in the Glutenfunen cohort had total IgA <0.5 g/L, the lowest value being 0.02 g/L. All were negative for TG2-IgA as well as for transglutaminase-2 IgG and endomysial IgG antibodies.
In conclusion, the prevalence of CeD in the Glutenfunen cohort was 2.6% (14 + 19)/1266 participants. The prevalence of CeD in girls was 2.9% (9 + 15)/811 participants, and in boys 2%, (5 + 4)/455 participants. The prevalence of Celiac autoimmunity was 3.2% (19 + 14 + 8)/1266. In the subsample of DNBC at Funen, a minimum prevalence was estimated to be 0.6% (14 + 14 + 19)/7431 participants, assuming that all undiagnosed CeD cases were participants of the Glutenfunen cohort.
Characteristics of Screen-detected Celiac Disease
Only a few differences were found in the basic characteristics across participants with and without diagnosed CeD in the Glutenfunen cohort (Table 2). Participants with diagnosed CeD were more often girls (79% vs 64%) and had slightly more gastrointestinal symptoms (42% vs 38%). They were slightly thinner but there were no signs of failure to thrive with regard to height or age at menarche. Participants with newly detected CeD had significantly lower low density lipoprotein (LDL)-cholesterol, 1.65 mmol/L compared with 1.99 mmol/L; however, there was no difference in total cholesterol or triglycerides. At the follow-up visit (3 and 9 months), all participants were still adhering to the gluten-free diet (Table 2).
TABLE 2 -
Characteristic of participants with newly detected celiac disease compared with participants in the Glutenfunen Cohort
||Detected CeD, N = 19
||The Glutenfunen Cohort, N = 1247
|Gender, females, n (%)†
|BMI, expressed in z score, (SD)
|Blood pressure systolic, expressed in z score (SD)
|Total IgA, g/L (SD)
|Total cholesterol, mmol/L (SD)
|LDL cholesterol, mmol/L (SD)
|HDL cholesterol, mmol/L (SD)
|Autoimmunity in first-degree relatives,‡ n (%)
|Date of menarche, year (SD)‡
|Number of gastrointestinal symptoms‡, n (%)
Ced = celiac disease; SD = standard deviation.
∗Values are mean for normally distributed variables, median and percentiles for nonnormally distributed variables and n (valid percent) for categorical variables. If the values differs significant P < 0.05 there are marked with an asterisk (∗).
†On the basis of the Danish national personal identification number (CPR).
‡On the basis of the questionnaire to the participants.
The main finding of this population-based cohort is the high prevalence of CeD, suggesting that CeD affects 1 in 40 of the adolescent population in Denmark. This makes CeD one of the most common autoimmune disease in the population. In addition, more than half of the CeD cases were asymptomatic or oligosymptomatic and were only discovered because of the screening procedure.
Prevalence of Celiac Disease in Denmark
We have earlier reported a prevalence of diagnosed CeD of 83 per 100,000 among Danish children and adolescents, which is much lower than the 2.6% prevalence in the Glutenfunen cohort. Even the minimum prevalence is 0.6%, which is in line with a 2015 Danish screening study of adults (8). The newly detected CeD participants in the Glutenfunen cohort were asymptomatic or oligosymptomatic, and thus, would not be recognized in the clinic. The prevalence of diagnosed CeD among nonparticipants was 0.2% based on data from the Danish National Patient Registry, indicating that those with known CeD were more eager to participate in the Glutenfunen cohort, which may lead to an overestimation of the prevalence of CeD in the Glutenfunen cohort.
Our results, however, indicate a true increase in prevalence of CeD in Denmark. After the Swedish celiac epidemic in infants (1984–1995) (9), a high prevalence of CeD among Swedish children of 2.2% (19) has consistently been reported. In a screening study of children with diabetes type 1 in both Denmark and the southern part of Sweden, an expected higher prevalence of celiac autoimmunity in Swedish children compared with Danish children was found. Remarkably, the prevalence of celiac autoimmunity in the Danish control group (healthy school children without known diabetes or CeD) was 2.8%, similar to the prevalence of CeD in the Glutenfunen cohort (20). The results could not be explained by differences in HLA genotype, which highlights the importance of environmental factors and that the prevalence of CeD in Denmark now resembles that of Sweden.
Screen-detected Celiac Disease
More than half (19/33) of the CeD cases were detected because of the screening procedure, despite Denmark's easy access to health care, a free health care system, and a generally high awareness of CeD among general practitioners because of several celiac research projects conducted on Funen over the last couple of years (21). At the clinical visit before endoscopy, more than half of the screen-detected CeD cases were asymptomatic, suggesting that the participants did not experience clinical symptoms worth mentioning to a health professional. Several studies have, however, demonstrated that undetected, and thus, untreated CeD poses a risk not only for morbidity but also mortality (4,5), thus highlighting the importance of further research into the pros and cons of celiac screening in early adulthood.
Most studies on CeD in adolescents have centered on psychological consequences and adhesion to a gluten-free diet (22) rather than on epidemiological data. In our cohort, the phenotype of CeD was mild with few symptoms, and we can only guess whether the symptoms would have become more severe later in life. A limitation of our study is that we were not able to identify when the seroconversion of Tg-2 antibodies occurred.
Interestingly, we found a significantly lower LDL in newly detected CeD participants compared with the Glutenfunen cohort and a tendency towards lower blood pressure and BMI. A similar pattern was observed in another screening study from Great Britain, which revealed a significantly lower level of LDL, diastolic blood pressure, and weight in EMA-positive participants compared with EMA-negative participants, neither of which group was on a gluten-free diet (23). These data indicate that undetected CeD may have a favorable cardiovascular risk profile, as previously suggested (24). This is, however, contrary to a recent study, which found an increased cardiovascular risk in undetected CeD (4) perhaps because of lower levels of high-density lipoprotein (HDL) in untreated CeD (25).
The 2 major strengths of this study are the prevalence estimates based on a population-based cohort and confirmed by duodenal biopsies. Four asymptomatic participants who refused gastroscopy were EMA-IgA negative and were very unlikely to have CeD (1). To avoid recall bias, we prospectively, before blood sampling, asked the participants for gastrointestinal symptoms, which allowed us to show that the majority (18 of 19) of the newly diagnosed CeD patients were asymptomatic or had mild subclinical symptoms.
As a limitation of our study, the cohort consists of only 17% of the original total DNBC Funen cohort, which could have introduced selection bias but no differences were observed when comparing age and postal code between participants and nonparticipants. Indeed, assuming that no undiagnosed CeD cases exist among nonparticipants in the DNBC cohort, the minimum prevalence is 0.6% (47 of 7431), which is still higher than previously reported among Danish children and adolescents (7).
Typical of similar studies, our study revealed a greater participation rate by girls (26). As female gender is associated with a higher risk of CeD, this may have led to an overestimation of the prevalence in the Glutenfunen cohort, the gender-specific prevalence being 2.9% for girls and 2% for boys (27). Although it is not clear why girls have a higher risk of CeD, a similar pattern can be observed in other autoimmune diseases (28).
Sixty percentage of the participants in the Glutenfunen cohort had at least 1 gastrointestinal symptom, which is in line with a large American study surveying 70,000 adults, where 61% reported more than 1 gastrointestinal symptom each week (29) and could indicate self-selection of individuals with symptoms. There was, however, no difference detected between participants with CeD and the seronegative Glutenfunen cohort. Furthermore, the majority of newly detected CeD patients were asymptomatic or monosymptomatic, indicating that self-selection does not have a substantial effect (30).
In conclusion, we found that 1 in 40 Danish adolescents may have CeD, indicating both underdiagnosis and perhaps a true increase in occurrence of the disease in a country where CeD has historically been rare. This work suggests the urgent need to understand the impact that this largely unrecognized disease has on society and why such an increase has occurred a decade or 2 after that of our close neighboring country, Sweden.
The authors would like to thank the DNBC participants, especially those who entered the Glutenfunen cohort.
1. Husby S, Koletzko S, Korponay-Szabo IR, et al. ESPGHAN Working Group on Coeliac Disease Diagnosis, ESPGHAN Gastroenterology Committee, European Society for Pediatric Gastroenterology, Hepatology, and Nutrition. European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease. J Pediatr Gastroenterol Nutr
2. Lebwohl B, Rubio-Tapia A. Epidemiology, presentation, and diagnosis of celiac disease. Gastroenterology
3. Walker MM, Murray JA, Ronkainen J, et al. Detection of celiac disease and lymphocytic enteropathy by parallel serology and histopathology in a population-based study. Gastroenterology
4. Karhus LL, Skaaby T, Petersen J, et al. Long-term consequences of undiagnosed celiac seropositivity. Am J Gastroenterol
5. Rubio-Tapia A, Kyle RA, Kaplan EL, et al. Increased prevalence and mortality in undiagnosed celiac disease. Gastroenterology
6. Weile B, Krasilnikoff PA. Extremely low incidence rates of celiac disease in the Danish population of children. J Clin Epidemiol
7. Dydensborg S, Toftedal P, Biaggi M, et al. Increasing prevalence of coeliac disease in Denmark: a linkage study combining national registries. Acta Paediat
8. Horwitz A, Skaaby T, Karhus LL, et al. Screening for celiac disease in Danish adults. Scand J Gastroenterol
9. Myleus A, Ivarsson A, Webb C, et al. Celiac Disease Revealed in 3% of Swedish 12-year-olds Born During an Epidemic. J Pediatr Gastroenterol Nutr
10. Ludvigsson JF, Murray JA. Epidemiology of celiac disease. Gastroenterol Clin North Am
11. Weile B, Grodzinsky E, Skogh T, et al. High prevalence rates of adult silent coeliac disease, as seen in Sweden, must be expected in Denmark. Apmis
12. Olsen J, Melbye M, Olsen SF, et al. The Danish National Birth Cohort--its background, structure and aim. Scand J Public Health
13. Henriksen DP, Rasmussen L, Hansen MR, et al. Comparison of the five Danish regions regarding demographic characteristics, healthcare utilization, and medication use--a descriptive cross-sectional study. PLoS One
14. Tinggaard J, Aksglaede L, Sorensen K, et al. The 2014 Danish references from birth to 20 years for height, weight and body mass index. Acta Paediatrica
15. National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics
2004; 114: (2 Suppl 4th Report): 555–576.
16. Oberhuber G. Histopathology of celiac disease. Biomed Pharmacother
17. Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scand J Public Health
2011; 39: (7 Suppl): 30–33.
18. Harris PA, Taylor R, Minor BL, et al. The REDCap consortium: building an international community of software platform partners. J Biomed Inform
19. Ivarsson A, Myleus A, Norstrom F, et al. Prevalence of childhood celiac disease and changes in infant feeding. Pediatrics
20. Adlercreutz EH, Svensson J, Hansen D, et al. Prevalence of celiac disease autoimmunity in children with type 1 diabetes: regional variations across the Oresund strait between Denmark and southernmost Sweden. Pediatr Diabetes
21. Toftedal P, Hansen DG, Nielsen C, et al. Questionnaire-based case finding of celiac disease in a population of 8-to 9-year-old children. Pediatrics
22. Ludvigsson JF, Agreus L, Ciacci C, et al. Transition from childhood to adulthood in coeliac disease: the Prague consensus report. Gut
23. West J, Logan RF, Hill PG, et al. Seroprevalence, correlates, and characteristics of undetected coeliac disease in England. Gut
24. Whorwell PJ, Alderson MR, Foster KJ, et al. Death from IschÆmic Heart-Disease and Malignancy in Adult Patients with CŒliac Disease. The Lancet
25. Brar P, Kwon GY, Holleran S, et al. Change in lipid profile in celiac disease: beneficial effect of gluten-free diet. AmJ Med
26. Tielemans MM, Jaspers Focks J, van Rossum LG, et al. Gastrointestinal symptoms are still prevalent and negatively impact health-related quality of life: a large cross-sectional population based study in the Netherlands. PLoS One
27. Wingren CJ, Agardh D, Merlo J. Sex differences in coeliac disease risk: a Swedish sibling design study. Digestive and Liver Disease
28. Ji J, Sundquist J, Sundquist K. Gender-specific incidence of autoimmune diseases from national registers. J Autoimmun
29. Almario CV, Ballal ML, Chey WD, et al. Burden of gastrointestinal symptoms in the United States: results of a nationally representative survey of over 71,000 Americans. Am J Gastroenterol
30. White LE, Bannerman E, McGrogan P, et al. Childhood coeliac disease diagnoses in Scotland 2009-2010: the SPSU project. Arch Dis Child