What Is Known
- Celiac disease presentation is changing.
- Occurrence of celiac disease has been variable between different regions.
What Is New
- Incidence and prevalence of celiac disease have continued to increase in children during the past 15 years in Olmsted County, Minnesota.
- Clinical and pathologic presentations of celiac disease are changing over time (more nonclassical and asymptomatic cases are emerging).
Celiac disease (CD) is one of the most common immune-mediated disorders in both children and adults (1). In patients with CD, consuming grain-containing gluten will result in characteristic histologic changes in the small bowel (2). CD primarily affects individuals of northern European descent, with an estimated prevalence of 1% to 3% in North America and Europe (3–7). According to the American College of Gastroenterology (ACG) and the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition (NASPGHAN), CD diagnostic guidelines in both adults and children require positive serologic markers followed by a confirmatory small bowel biopsy (SBB) with characteristic histologic features (2,8). The most recent guidelines from the European Society for Pediatric Gastroenterology Hepatology and Nutrition (ESPGHAN) suggested the possibility of omitting the confirmatory SBB in children who meet specific criteria, particularly those with symptoms of CD (9).
The occurrence of CD varies widely across the literature (10). Based on many epidemiologic studies, including the data from Olmsted County, Minnesota, in the United States, prevalence and incidence of CD increased considerably between 1950 and 2001 (11–13). It is not clear whether this was a true increase in occurrence or was secondary to the availability of more sensitive testing such as antitissue transglutaminase immunoglobulin A antibodies, or was due to increased awareness and screening of CD. Recent reports from Finland and Italy suggest that the previously observed increments in CD incidence in children has plateaued, whereas other reports from Scotland, the Netherlands, and the United Kingdom suggest that the incidence is still increasing (4,12,14–17).
It is known that symptomatic patients with CD only represent the tip or the celiac iceberg (18). CD presents with different clinical manifestations at different ages. Diarrhea and failure to thrive are common in infants and young children, and extraintestinal manifestations are more common in older children and adolescents (19). The clinical presentation of CD in children and adolescents has been divided into classical and nonclassical presentation. Classical CD usually presents with signs and symptoms of malabsorption, such as diarrhea and poor growth (20), whereas nonclassical CD presents with nonspecific gastrointestinal (GI) symptoms (eg, nonspecific abdominal pain, nausea, vomiting, and constipation) and/or extra-GI manifestations (eg, oral ulcers, hepatitis, dermatitis herpetiformis, iron deficiency anemia, and unexplained bone fractures) (2,21–23). Over the past decade, more children are presenting with nonclassical presentation of CD (20,24). Furthermore, recent data suggest that up to 19% of children with a new CD diagnosis are overweight or obese (25).
The majority of patients with CD are still undiagnosed, despite increased awareness and the availability of sensitive serologic testing (26). In an effort to identify asymptomatic patients with CD, many high-risk patients, such as those with a family history of CD, diabetes mellitus type 1 (DM1), thyroid disease, immunoglobulin (IgA) deficiency, and Down syndrome, are being screened regardless of their symptoms (27).
We performed a retrospective cohort study aimed at examining the trends in incidence and prevalence of childhood CD in Olmsted County, Minnesota. We assessed the change in clinical and histologic presentation of CD during the past 14 years.
We conducted our study in Olmsted County, which is located in southeastern Minnesota, in the United States. The Olmsted County population has the same age, sex, and ethnic characteristics of Minnesota in general, with a majority of non-Hispanic whites (82%) and small minorities of African American (6%), Asian (6%), Hispanic (5%), and mixed ethnicities (1%) (28). Olmsted County provides an exceptional location to perform population-based studies such as this because almost all Olmsted County, MN residents, their all inpatient and outpatient medical events, and health care providers are linked and retrieved for research under the auspices of the Rochester Epidemiology Project (REP) starting in 1966 (29,30). Since then, the REP has been continuously supported by a National Institutes of Health, which provides a continuously updated database from Olmsted County's 2 main health care providers: Mayo Clinic and Olmsted Medical Center (29,30). We used the REP database to search for potential CD cases using International Classification of Diseases, Ninth Revision code: 579.0. We also screened the databases of previous studies conducted in Olmsted County for potential cases (13,31).
Case Ascertainment and Characterization
After obtaining approval from the institutional review boards of Mayo Clinic and Olmsted Medical Center, we retrospectively reviewed the records of Mayo Clinic and Olmsted Medical Center for our cohort of potential cases. The index date was defined as the date of clinical diagnosis. We identified all patients ages 18 years or younger with an index date between January 1, 1994 and December 31, 2014, who met the NASPGHAN or ESPGHAN guidelines for CD diagnosis (2,8,9). Incidence was calculated between 2000 and 2014, and charts from 1994 to 1999 were reviewed only to calculate prevalence. The following data were collected from the charts: birthdate, race, sex, date of diagnosis, family history of CD, body mass index (BMI) at index date, clinical presentations at the time of diagnosis, CD-specific antibodies tests, and IgA levels. Pathology reports of SBB were reviewed by 1 author (E.A.). The degree of mucosal injury on the SBB was classified as normal, increase in intraepithelial lymphocytes (IELs) without villous atrophy, partial villous atrophy, or complete villous atrophy (32). CD cases were categorized into 3 groups based on the presenting symptoms at the time of diagnosis. The first group included children with symptoms of malabsorption (classical CD). The second group included children with other nonspecific GI or extra-GI symptoms (nonclassical CD). The third group included asymptomatic children diagnosed based on screening of high-risk groups for CD (20).
Patients’ growth parameters were reported following the current guidelines that recommend assessing body weight percentiles in children according to their age and sex. For children younger than 24 months, we used World Health Organization standards to determine the weight for length (W/L) percentile (33). For children 25 months or older, we used the Centers for Disease Control and Prevention growth charts to get the BMI percentile (34).
Inclusion criteria include positive CD serology markers such antitissue transglutaminase immunoglobulin A antibodies, antiendomysial antibodies, or deamidated gliadin peptide (DGP) antibodies; confirmatory SBB with CD characteristic histologic findings (eg, increase in IEL, villous atrophy, and crypts hyperplasia) (32) or no SBB but meets the ESPGHAN criteria for avoiding the SBB; and authorization for using medical records for research.
Many CD cases in the present study were diagnosed by the principal investigator (I.A.). Olmsted County residency was confirmed electronically using the REP database. Cases were considered incident if the patient was living in Olmsted County at the time of diagnosis to exclude referral cases seeking tertiary care. For prevalence cases, we considered all patients (even those diagnosed elsewhere) and verified their place of residency and their age both on January 1, 2010 and December 31, 2014.
Patients who were older than 18 years at time of diagnosis, had negative SBB, or did not have confirmatory SBB without meeting ESPGHAN criteria for omitting SBB were excluded. Subjects who declined permission for use of their medical records for research purposes were excluded from the present study as required by Minnesota state law (Supplemental Digital Content 1, Fig., http://links.lww.com/MPG/A896) (35).
Descriptive statistics for categorical variables are reported as number (percentage) and continuous variables as median (interquartile range). Age-, sex-, and calendar-year–specific incidence rates of CD diagnosed with biopsy (or ESPGHAN criteria) were calculated by using the number of new pediatric patients (18 years or older) who were living in Olmsted County at the time of diagnosis as the numerator, whereas the denominator was based on REP census data, assuming that all children were at risk. Incidence rates were standardized to the 2010 US white population. The 95% confidence intervals (CIs) for incidence rates assumed the number of cases followed a Poisson distribution. The association of age, sex, and calendar year were accessed with multivariable Poisson regression. The functional form of calendar year was tested with linear trend contrasts, and the functional form of age was tested with Loess smoothing plots. Incidence rates were calculated by sex for 3 age categories (0–5, 6–14, and 15–18), with 3-year intervals. Generalized linear models (cumulative and binary logistic regression where appropriate) were used to test associations of presenting symptoms and villous atrophy with age, sex, and calendar year. Chi-square and Fisher exact tests were used for the associations of presenting symptoms, villous atrophy, and BMI where appropriate. Data were analyzed using SAS statistical software version 9.4 (SAS Institute Inc, Cary, NC). An alpha level of 0.05 was used as the level of statistical significance.
We identified a total of 102 confirmed CD patients who met our inclusion criteria (100 underwent SBB and 2 were diagnosed without SBB according to the new ESPGHAN guidelines) (9). Two patients were diagnosed outside of Olmsted County but subsequently moved to Olmsted at the time of our prevalence dates. Therefore, there were 100 incident cases between 2000 and 2014. Demographic characteristics of our incidence cohort are shown in Table 1. At the time of diagnosis, 23 (%) children were 5 years or younger, 65 (%) were between 6 and 14 years, and 12 (%) were 15 years or older with 59 (%) female patients. About half (49%) of the incident cases included in the present study were diagnosed in the last 5 years of the study, between 2010 and 2014. Only 20 (%) of the included incidence cases had their DGP antibodies checked and 26 (%) had the CD-associated human leukocyte antigen status checked.
Temporal Trend of Incidence
In the period from 2000 through 2014, the overall annual age- and sex-adjusted standardized incidence rate of CD was 17.4 per 100,000 person-years (95% CI, 14.0–20.8). Between 2000 and 2002, the annual adjusted incidence rate was 8.2 per 100,000 person-years (95% CI, 2.8–13.3), whereas the incidence rate rose to its highest peak between 2009 and 2011, with 25.7 per 100,000 person-years (95% CI, 16.4–34.9). The increase in incidence appeared to be linear over time (P = 0.004) (Fig. 1), with no interactions with age or sex. The functional form of age was observed to be quadratic because there was high incidence in 6 to 14 year olds but a lower incidence for toddlers and high school–aged patients (Fig. 2). When added to a Poisson model including sex and linear calendar year, quadratic age was significant (P < 0.001). Sex was also a significant predictor of the number of incidence cases, with girls having a higher incidence (P = 0.04) (Supplemental Digital Content 2, Fig., http://links.lww.com/MPG/A897). Detailed incidence information is presented in Table 2.
Prevalence of childhood CD in Olmsted County was calculated at 2 points, January 1, 2010 and December 31, 2014. Thirty-nine children with confirmed CD were living in Olmsted County on January 1, 2010, and the age- and sex-adjusted prevalence was 103.6 per 100,000 person-years (95% CI 71.0–136.2). Prevalence increased in the following 5 years, with 65 cases in Olmsted County and an age- and sex-adjusted prevalence rate of 173.9 per 100,000 person-years (95% CI 131.4–216.4) on December 31, 2014, although this was not statistically significantly (P = 0.07). The highest prevalence rate reported in 2010 for boys was in the 15- to 18–year-old age group, yet for girls the highest prevalence was in the 6- to 14-year-old age group. In 2014 the highest prevalence rate reported across sexes was, however, in the high school–age group (15–18 years) (411.1 and 189.8 per 100,000 person-years, respectively), as seen in Supplemental Digital Content 3, Table, http://links.lww.com/MPG/A898.
The most common presenting symptoms in all the 100 incident cases were nonspecific GI or extra-GI symptoms (nonclassical CD) in 43 (43%) of included patients (Supplemental Digital Content 4, Fig., http://links.lww.com/MPG/A899). Approximately one-third of patients (34%) presented with diarrhea and/or weight loss (classical CD). The remaining 23 patients (23%) were asymptomatic and were diagnosed as part of our general practice of screening high-risk patient (family members, trisomy 21, Turner, or insulin-dependent diabetes) for CD (27). Sixteen of asymptomatic patients were screened because of family history of CD and 7 because of personal history of DM1. The mean age at diagnosis for patients presenting with classical CD was 8.09 years, and the mean age for children who had nonclassical CD or were asymptomatic was 10.19 years (P = 0.025). The most common nonspecific GI symptom was abdominal pain, which was reported in 37 patients, followed by other GI symptoms (Supplemental Digital Content 5, Fig., http://links.lww.com/MPG/A900). Nonclassical CD increased significantly over the years of our study (P = 0.004), whereas both silent CD and classical CD decreased, although classical presentation was not significant (P = 0.023 and P = 0.32, respectively). Younger children tended to have classical CD (P = 0.016), which can explain their younger age at diagnosis. Only 2 patients had villous atrophy confined to the first portion of the duodenum (Ultrashort CD) and presented with mild nonclassic symptoms (36).
Body Weight Assessment
The median BMI percentile was 51 (11–79). Sixty-seven incident cases (67%) had a normal BMI at diagnosis for their age and sex. Three patients (3%) were obese (BMI ≥95th percentile) and 14 (14%) were overweight (BMI ≥85th percentile) at the time of diagnosis. Only 9 (9%) were underweight (BMI <5th percentile). Weight records for 7 (7%) cases were missing. BMI (as categorized by standard definition (33,34)) was associated with clinical presentation. Specifically, patients with asymptomatic CD were more likely to be overweight or obese than patients with presenting symptoms (P = 0.045).
There was no correlation between the degree of villous atrophy on SBB and the type of clinical presentation (P = 0.43), which agrees with reports on a lack of association between severity of mucosal damage and clinical presentation (37). The degree of villous atrophy on the SBB was not found to be significantly related to BMI (P = 0.80), age, sex, or calendar year.
In the present study, we report the epidemiologic characteristics of childhood CD in Olmsted County, Minnesota. The diagnoses were made according to the current North American and European criteria (8,9). Unlike the previous epidemiologic studies in Olmsted County, we focused mainly on the pediatric population in this unique community. In the period from 2000 through 2014, 100 children were diagnosed with CD, with a total incidence of 17.4 per 100,000 person-years. We found a significant linear increase in total incidence. This finding is consistent with recent studies from Scotland, the Netherlands, and the UK (4,14–16). It also shows that the previously noticed increase in CD incidence in all ages in Olmsted County (13,31), from 1950 to 2001 and 2001 to 2010, has continued in children through 2014. The prevalence of CD in children has increased from 0.10% in 2010 to 0.17% in 2014. This prevalence is one of the highest in the current literature in this age category (4).
The underlying etiology for the increase in CD occurrence is unknown. Similar increase has been reported in other autoimmune diseases in children, such as DM1 which suggests the possibility of true increase in CD occurrence (38). On the contrary, the availability of sensitive serologic tests, screening of high risk patients, and the increased awareness of CD between medical providers and families may play a role in the apparent increase in the incidence and prevalence.
Based on the serology screening tests, the National Health and Nutrition Examination Survey shows that the national CD prevalence in children between 6 and 18 years in the United States was 0.59% in 2011 and 2012 (39). Compared to our result in Olmsted County, we think there is a gap between the prevalence of clinically diagnosed CD and the prevalence of CD found by screening tests. As a result, many silent and nonclassical CD cases are still undiagnosed despite the presence of excellent health care and a high awareness of CD in the small community of Olmsted County.
Ninety-eight (98%) of our cases met the NASPGHAN criteria. Only 2 patients (2%) were diagnosed without SBB according to the new ESPGHAN guidelines. Out of the 15 patients who were diagnosed after the ESPGHAN guidelines were published in 2012, SBB was avoided in only 1 patient (the other patient diagnosed without SBB was in 2008). As a result, most pediatricians in Olmsted County are still using SSB as the criterion standard for CD diagnosis. New diagnostic tests have been introduced over the past few years, including the DGP antibodies as new serologic markers and the assessment of CD-associated human leukocyte antigen status in difficult cases (40–42). These tests were performed in small number of our included cases that could be due to the tests being relatively new or being performed in difficult cases only.
According to a previous study in Olmsted County before 2001, all patients with CD were symptomatic, and the majority presented with classical malabsorptive symptoms of CD (13). After 2000, two-thirds of patients with CD presented with nonclassical symptoms or were diagnosed by screening of high-risk patients with DM1 or family history of CD. Moreover, only 9% of included cases were underweight at the time of diagnosis. These results affirm the previous studies that suggest a change in the classical malabsorptive picture of CD in children to a more subtle presentation. These studies suggest that more children with CD are presenting with nonclassical symptoms, and up to 19% are overweight or obese (25,43,44).
In a previous study in Olmsted County between 1950 and 2001, most patients had complete villous atrophy on the SBB (13). In our study, however, only 37% of our pediatric cohort had complete villous atrophy, and most patients had partial villous atrophy or only an increase in IEL in the period from 2000 through 2014. This trend can be the result of increased awareness of CD and the availability of sensitive tests resulting in earlier detection and diagnosis.
Limitations and Strengths
Our study has a few limitations, including the inherent limitations as a retrospective study, the small number of incident cases despite the population-based study, and the demographic of Olmsted County, in which >80% of the population were white. It is also possible that some patients started a gluten-free diet without medical intervention or evaluation. Our study has a few important strengths. First, our study was a population-based study. Second, our study setting has unique epidemiological advantages for retrospective population-based studies including self-contained health care environment and medical record linkage system through the REP for the almost entire Olmsted County population. Lastly, our case ascertainment was based on clinical criteria and pathology.
CD incidence and prevalence continue to increase in Olmsted County, Minnesota. Clinical presentation and epidemiological profile of CD are changing, specifically presenting with nonclassic milder clinical and pathological findings. Large properly designed population-based prospective studies are needed, to ascertain the actual changes in the incidence and clinical presentation of CD.
1. Catassi C, Kryszak D, Bhatti B, et al. Natural history of celiac
disease autoimmunity in a USA cohort followed since 1974. Ann Med
2. Rubio-Tapia A, Hill ID, Kelly CP, et al. ACG clinical guidelines: diagnosis and management of celiac
disease. Am J Gastroenterol
3. Green PH. Where are all those patients with celiac
disease? Am J Gastroenterol
4. Kang JY, Kang AH, Green A, et al. Systematic review: worldwide variation in the frequency of coeliac disease and changes over time. Aliment Pharmacol Ther
5. Hill ID. Celiac
disease—a never-ending story? J Pediatr
6. Ji J, Ludvigsson JF, Sundquist K, et al. Incidence
disease among second-generation immigrants and adoptees from abroad in Sweden: evidence for ethnic differences in susceptibility. Scand J Gastroenterol
7. 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
8. Hill ID, Dirks MH, Liptak GS, et al. Guideline for the diagnosis and treatment of celiac
disease in children: recommendations of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition. J Pediatr Gastroenterol Nutr
9. 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
10. Gujral N, Freeman HJ, Thomson ABR. Celiac
, diagnosis, pathogenesis and treatment. World J Gastroenterol
11. Choung RS, Ditah IC, Nadeau AM, et al. Trends and racial/ethnic disparities in gluten-sensitive problems in the United States: findings from the National Health and Nutrition Examination Surveys from 1988 to 2012. Am J Gastroenterol
12. Kivela L, Kaukinen K, Lähdeaho ML, et al. Presentation of celiac
disease in Finnish children is no longer changing: a 50-year perspective. J Pediatr
13. Murray JA, Van Dyke C, Plevak MF, et al. Trends in the identification and clinical features of celiac
disease in a North American community, 1950–2001. Clin Gastroenterol Hepatol
14. Angeli G, Pasquini R, Panella V, et al. An epidemiologic survey of celiac
disease in the Terni area (Umbria, Italy) in 2002–2010. J Prev Med Hyg
15. Burger JPW, Roovers EA, Drenth JP, et al. Rising incidence
disease in the Netherlands; an analysis of temporal trends from 1995 to 2010. Scand J Gastroenterol
16. White LE, Merrick VM, Bannerman E, et al. The rising incidence
disease in Scotland. Pediatrics
17. West J, Fleming KM, Tata LJ, et al. Incidence
disease and dermatitis herpetiformis in the UK over two decades: population-based study. Am J Gastroenterol
18. Catassi C, Fabiani E, Ratsch IM, et al. The coeliac iceberg in Italy. A multicentre antigliadin antibodies screening for coeliac disease in school-age subjects. Acta Paediatr
19. D’Amico MA, Holmes J, Stavropoulos SN, et al. Presentation of pediatric celiac
disease in the United States: prominent effect of breastfeeding. Clin Pediatr
20. Ludvigsson JF, Leffler DA, Bai JC, et al. The Oslo definitions for coeliac disease and related terms. Gut
21. Green PH. The many faces of celiac
disease: clinical presentation of celiac
disease in the adult population. Gastroenterology
2005; 128 (4 suppl 1):S74–S78.
22. Kavak US, Yuce A, Kocak N, et al. Bone mineral density in children with untreated and treated celiac
disease. J Pediatr Gastroenterol Nutr
23. Mora S, Weber G, Barera G, et al. Effect of gluten-free diet on bone mineral content in growing patients with celiac
disease. Am J Clin Nutr
24. Tapsas D, Hollen E, Stenhammer L, et al. The clinical presentation of coeliac disease in 1030 Swedish children: changing features over the past four decades. Dig Liver Dis
25. Reilly NR, Aguilar K, Hassid BG, et al. Celiac
disease in normal-weight and overweight children: clinical features and growth outcomes following a gluten-free diet. J Pediatr Gastroenterol Nutr
26. Nenna R, Tiberti C, Bonamico M, et al. The celiac
iceberg: characterization of the disease in primary schoolchildren. J Pediatr Gastroenterol Nutr
27. Fasano A, Berti I, Gerarguzzi T, et al. Prevalence
disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med
28. St Sauver JL, Grossardt BR, Leibson CL, et al. Generalizability of epidemiological findings and public health decisions: an illustration from the Rochester Epidemiology
Project. Mayo Clin Proc
29. Rocca WA, Yawn BP, St Sauver JL, et al. History of the Rochester Epidemiology
Project: half a century of medical records linkage in a US population. Mayo Clin Proc
30. Melton LJ 3rd. History of the Rochester Epidemiology
Project. Mayo Clin Proc
31. Ludvigsson JF, Rubio-Tapia A, Van Dyke CT, et al. Increasing incidence
disease in a North American population. Am J Gastroenterol
32. Oberhuber G. Histopathology of celiac
disease. Biomed Pharmacother
33. Centers for Disease Control and Prevention. WHO growth standards are recommended for use in the U.S. for infants and children 0 to 2 years of age. September 09, 2010 [cited March 01, 2016]. http://www.cdc.gov/growthcharts/who_charts.htm
. Accessed October 15, 2016.
35. Melton LJ 3rd. The threat to medical-records research. New Engl J Med
36. Mooney PD, Kurien M, Evans KE, et al. Clinical and immunologic features of ultra-short celiac
37. Marsh MN, Crowe PT. Morphology of the mucosal lesion in gluten sensitivity. Baillieres Clin Gastroenterol
38. Lipman TH, Levitt Katz LE, Ratcliffe SJ, et al. Increasing incidence
of type 1 diabetes in youth: twenty years of the Philadelphia Pediatric Diabetes Registry. Diabetes Care
39. Rubio-Tapia A, Ludvigsson JF, Brantner TL, et al. The prevalence
disease in the United States. Am J Gastroenterol
40. Sugai E, Vázquez H, Nachman F, et al. Accuracy of testing for antibodies to synthetic gliadin–related peptides in celiac
disease. Clin Gastroenterol Hepatol
41. Amarri S, Alvisi P, De Giorgio R, et al. Antibodies to deamidated gliadin peptides: an accurate predictor of coeliac disease in infancy. J Clin Immunol
42. Pallav K, Kabbani T, Tarig S, et al. Clinical utility of celiac
disease-associated HLA testing. Dig Dis Sci
43. Ravikumara M, Tuthill DP, Jenkins HR. The changing clinical presentation of coeliac disease. Arch Dis Child
44. Diamanti A, Capriati T, Basso MS, et al. Celiac
disease and overweight in children: an update. Nutrients