PEDIATRICS Vol. 107 No. 1 January 2001, pp. 42-45

Serological Screening for Celiac Disease in Healthy 2.5-Year-Old Children in Sweden

Anneli K. Carlsson, MD^*, Irene E. M. Axelsson, MD, PhD^*, Stefan K. Borulf, MD, PhD^*, Anders C. A. Bredberg, MD, PhD, and Sten-A. Ivarsson, MD, PhD^*

From the Departments of ^* Pediatrics and  Medical Microbiology, University of Lund, University Hospital, Malmö, Sweden.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Objective.  The study was designed to investigate the prevalence of celiac disease (CD) among 2.5-year-old children in a Swedish urban population with a high incidence of CD.

Material and Methods.  Six hundred ninety apparently healthy children, born in the 12-month period of July 1992 through June 1993, were screened for immunoglobulin A (IgA) antigliadin antibodies and IgA antiendomysium antibodies, and those antibody-positive at repeated testing were further investigated with intestinal biopsy.

Results.  Of the 690 children, 6 were both IgA antigliadin antibody- and IgA antiendomysium antibody-positive, and 7 were antiendomysium antibody-positive but antigliadin antibody-negative. Jejunal biopsy, performed in 12 cases, manifested partial or total villous atrophy in 8 cases. Thus, together with an additional child whose parents declined the offered biopsy, but whose response to a gluten-free diet confirmed the presence of CD, the prevalence of CD in the study series was 1.3% (9/690; 95% confidence interval: .4-2.2). However, independent of the study, an additional 22 cases of symptomatic, biopsy-verified CD have already been detected in the birth cohort of 3004 children.

Conclusions.  The prevalence of CD in our study series was high, at least 1.0%, but may be as high as 2.0% if the frequency of silent CD is as high as we have found in the remaining unscreened cohort. These findings confirm that CD is one of the most common chronic disorders.  Key words:  celiac disease, immunoglobulin A antigliadin antibodies, immunoglobulin A antiendomysium antibodies.

The incidence of celiac disease (CD) among Swedish children has increased from 0.031% to 0.29% during recent decades.^1,2 This is in contrast to other countries, for which decreasing^3,4 or unchanging⁵ incidence rates have been reported. In Sweden, high-incidence rates have been reported for young children born after 1982,^1,2 when infant-feeding patterns were changed; the introduction of gluten was delayed from 4 months to 6 months of age, but its intake was increased.⁶

Changing patterns in CD, from the classic presentation in infants to mild and atypical signs later in childhood and adolescence, have been reported from other countries.^7,8 The atypical and mild presentation of CD has also been reported by our neighboring countries but differs from ours in which the predominant clinical picture is a child with florid symptoms, diagnosed before 2 years of age.^9,10

Many CD screening studies have been performed in recent years.^11-13 Obviously screening will detect undiagnosed cases of CD, with atypical symptoms or even without symptoms despite typical enteropathy in the jejunal mucosa. High prevalences of CD have been detected in family studies^14,15 and among children with Down syndrome¹⁶ and among patients with type 1 diabetes mellitus.^17,18

One view of the epidemiology of CD is the celiac iceberg theory in which the undiagnosed cases are seen as constituting the bulk of the iceberg below the waterline and the diagnosed, symptomatic cases are seen as constituting its tip, visible above the surface.¹⁹ This model can also be used to interpret the high incidence among infants in Sweden. Our higher and more abrupt gluten introduction during infancy, compared with other countries, might provoke susceptible individuals to develop symptoms early in life, thus rendering more of the iceberg tip visible.

This study was designed to investigate the prevalence of children with undiagnosed CD in a population with a high incidence of CD.

	METHODS

Top Abstract Methods Results Discussion Conclusion References

Patients

Malmö, a Swedish city with a population of 240 000, is served by a single tertiary level hospital and a single Department of Pediatrics. Malmö has a comprehensive child health service, where 98% of all infants attend the child health centers regularly.²⁰ All of the infants with CD born in Malmö during the 12-month period (July 1992 through June 1993) were identified. The diagnostic criteria used were those recommended by the European Society for Pediatric Gastroenterology and Nutrition.²¹ From the same birth cohort (n = 3004), 1287 healthy children attending their local child health center for the routine 2.5-year-old health control were invited to participate in a CD screening study. Of the 992 children whose parents were initially interested in joining the study, 690 (359 girls and 331 boys) actually attended for the blood sampling. No attempt has been made to analyze the dropout group. The blood samples were collected at a mean age of 2 years 8 months (range: 2 years 3 months to 3 years 5 months).

The screening procedure was capillary blood analysis for both IgA (immunoglobulin A) antigliadin antibodies (AGAs) and IgA antiendomysium antibodies (EMAs). Children with normal levels of AGA and EMA were considered not to have CD and took no further part in the study, whereas those who were EMA- or AGA-positive underwent further examination by a gastroenterologist and a repeat test for EMA and AGA. Those AGA- or EMA-positive at repeat testing were offered jejunal biopsy (Fig 1).

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Diagnostic procedures.

The study design was approved by the Human Research Ethics Committee of the Faculty of Medicine, University of Lund.

Laboratory Methods

AGAs were measured with a commercial microplate enzyme-linked immunosorbent assay kit (Gluten IgA EIA, Pharmacia, Uppsala, Sweden). Microplate strips were coated with gliadin. Patient serum was diluted 1/200. IgAs were detected using galactosidase-conjugated rabbit antihuman IgA and the chromogenic substrate nitrophenyl--galactoside, the amount of specific IgA being proportional to the absorbance measured at 405 nm. As recommended by the manufacturer, results were expressed in arbitrary units (AUs), each unit representing 1% of the optical density value of the positive control.

The cutoff level for a positive test result was set at 25 AU, for which, in children <5 years of age, the sensitivity of the method has been shown to be 92% and the specificity 84% (as shown by the manufacturer's analysis of sera from 85 children <5 years of age with CD and 144 without CD).

EMAs were tested with indirect immunofluorescence analysis using commercially available fixed sections of monkey esophagus (distal third part) (BioSystems, Barcelona, Spain) as the antigen substrate. Patient serum was diluted 1/5 in phosphate-buffered .15 M NaCl (pH 7.6), 1% bovine serum albumin. IgAs specific for endomysium were detected with a fluorescein isothiocyanate-labeled antihuman IgA conjugate (BioSystems). The result was expressed as the highest dilution factor giving a positive fluorescence pattern. All sera manifesting fluorescence (titer: >5) were considered to be positive. The great majority of all EMA studies have shown this serologic marker to be associated with a very high specificity (98%-100%) and a somewhat lower sensitivity (78%-100%) in series of pediatric patients from North America²² or Sweden.²³ In one of these studies,²⁴ the sensitivity in children <5 years of age was found to be similar to that in children 5 years of age or older (95% vs 100%).

We routinely include in-house negative and positive controls for both AGA and EMA in every analysis. Moreover, for these 2 tests our laboratory takes part in the Swedish National and the United Kingdom National External Quality Assessment Schemes.

Biopsy

A small bowel biopsy was performed with a Watson (Ferrari Medical Ltd, Middlesex, England) capsule under fluoroscopic control at the level of the ligament of Treitz. The specimens were immersed in formaldehyde solution and examined histologically at the Department of Pathology, University Hospital, Malmö. The intestinal mucosa was classified as normal, subnormal (villous length/crypt length <2, increased number of inflammatory cells in the mucosa with or without damage to the surface epithelium and brush border), or total villous atrophy (flat mucosa). The revised criteria for the diagnosis of CD were used.²¹

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

Of the birth cohort as a whole (n = 3004) and independently of the study, 22 children (.7% or 1/143) had already had clinical CD verified by biopsy before selection of the study series.

EMA-Positive

In the screening study itself, 1.9% (13/690; 95% confidence interval [CI]:.1-2.9) of the children tested were found to be EMA-positive (6 of them being AGA-positive as well). Three of the 13 EMA-positive children identified at screening (with titers of 1/5 in 2 cases and of 1/40 in the third child) were EMA-negative at retesting; none of these children was AGA-positive.

AGA-Positive

AGA testing showed 4.1% (28/690; 95% CI: 2.6-5.6) of the children to be AGA-positive (6 of them being EMA-positive as well, as mentioned above). Of the 22 AGA-positive/EMA-negative children, 10 (45%) were AGA-negative at retesting. An additional 7 children manifested decreases in AGA levels at retesting, the values ranging from 25 to 30 AUs (the cutoff level for AGA-positivity being 25 AU). Because these children were also asymptomatic, their parents declined the offer of jejunal biopsy. Two children, 1 with an AGA level of 25 AU and the other with a level of 42 AU, did not attend for retesting despite reminders.

Biopsy Findings

Of the 10 children EMA-positive at retesting, 9 underwent jejunal biopsy; the parents of the reminding child (EMA titer of 320) declined the biopsy offer. Jejunal biopsy was also performed in 3 AGA-positive/EMA-negative children. Of the 12 biopsies performed, 8 were characteristic of CD with villous atrophy. All 8 children (5 girls and 3 boys) were EMA-positive, but 3 of those children were AGA-negative. Of the 4 children with normal biopsy, 1 was EMA-positive/AGA-negative (EMA titer of 1/40); the remaining 3 were AGA-positive/EMA-negative, 2 of them having cow's milk allergy and the third child, a Giardia lamblia infection.

Outcome at Subsequent Follow-Up

All 8 children with biopsy-verified CD reported above were put on a gluten-free diet. In 3 cases, the symptoms (eg, nightmares, abdominal pain, and diarrhea) disappeared and 1 child also gained 1 standard deviation (SD) in weight. Of the 5 remaining children, all of who were asymptomatic, 3 gained 1 SD in weight and 1 gained .5 SD in both weight and height. In addition, 1 girl (whose parents declined the biopsy offer) was both EMA- and AGA-positive (EMA titer: 1/320; AGA level: 30 AU) and suffered from abdominal pain. On introduction of a gluten-free diet, her antibody levels normalized, the pain disappeared, and she gained .5 SD in height and 1 SD in weight, an outcome that confirms the diagnosis of CD in this case.

Thus, 9 of the 690 children in our screening series were considered to have silent CD, a prevalence of 1.3% (95% CI: .4-2.2) at 2.5 years of age. In addition as mentioned above, CD had already been diagnosed in an additional .7% (22/3004) of the birth cohort, based on clinical symptoms, positive serology, and biopsy, which verified the diagnoses.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

In this screening study, 13 children were found to be EMA-positive, of whom nine were biopsied, all but one biopsy manifesting jejunal changes characteristic of CD.

Twenty-two children were AGA-positive/EMA-negative, of whom 3 with persistently high levels of AGA (43-95AU) had normal biopsies, the AGA levels being attributable to other causes. In addition to the 8 children with biopsy-verified CD, a girl with high AGA and EMA levels and abdominal pain, but whose parents declined biopsy offer, was confirmed to be a case of CD by her satisfactory response to a gluten-free diet. Thus, the prevalence of CD in the screening series was 1.3% (95% CI: .4-2.2).

Moreover, independent of and before the screening study, an additional 22 cases had been detected in the birth cohort as a whole, already then suggesting a minimum of .7% (22/3004), a figure higher than the incidence of CD among 2-year-olds reported in another recent Swedish study, which was reported to have increased from .03 (1/3000) to .29 (1/340) over a 20-year period.¹ Together, the prevalence of 1.3% (95% CI: .4-2.2) in the screened group and the 22 known children in the cohort, the total prevalence may be between 1.0% (31/3004) and 2.0% (1.3% of the remaining unscreened population in this cohort, ie, 2292). The prevalence is high compared with data from other screening studies.^11,24

It has been hypothesized that most children with CD are symptomatic early in life because of the high gluten content of infant diets.⁹ However, the present finding of high prevalence of both symptomatic and asymptomatic silent CD cases suggests that this interpretation may be open to question and that, instead, the standard gluten-rich diet may induce asymptomatic disease in children mildly predisposed to CD and more aggressive, symptomatic disease in highly susceptible children. However, more unlikely, they may represent a subgroup consuming a lower amount of gluten spontaneously.

Another question raised by the present findings is whether screening for silent CD is warranted. Of the 9 children of CD detected in our screening, 4 were characterized by the presence of symptoms, such as abdominal pain, diarrhea, and proneness to nightmares, the relationship of which to CD was borne out by disappearance when the children were put on a gluten-free diet. The remaining 5 children, who were asymptomatic, gained in weight and/or height when put on a gluten-free diet. The findings suggest that quality of life was enhanced for these children by having their CD detected. It is, of course, possible that it was only a matter of time before their CD would have been detected anyway from the emergence of symptoms.

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

In summary, the prevalence of CD in our series is, except for the study conducted by Catassi et al²⁵ in the Sahara, the highest yet reported. The high gluten intake among very young Swedish children may increase the prevalence of CD. Whether it also induces CD among susceptible children is unknown. The recommended age for the introduction of gluten was changed from 6 to 4 months in 1996. The effect of this in terms of the development of CD remains to be investigated.

	ACKNOWLEDGMENTS

This study was supported by grants from Medical Faculty, University of Lund, Health Services Administration, Malmö, Swedish Coeliaci Association, the Sven Jerring Fund, Swedish Medical Research Council Project K97-27X-12 274.

	FOOTNOTES

Received for publication Nov 18, 1999; accepted May 24, 2000.

Reprint requests to (A.K.C.) Department of Pediatrics, University Hospital, Malmö, 205 02, Malmö, Sweden. E-mail: anneli.k.carlsson{at}skane.se

	ABBREVIATIONS

CD, celiac disease; IgA, immunoglobulin A; AGA, IgA antigliadin antibody; EMA, IgA antiendomysium antibody; AU, arbitrary units; CI, confidence interval; SD, standard deviation.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. Ascher H, Krantz I, Kristiansson B Increasing incidence of coeliac disease in Sweden. Arch Dis Child 1991; 66:608-611 [Abstract/Free Full Text]
2. Cavell B, Stenhammar L, Ascher H, Increasing incidence of childhood coeliac disease in Sweden: results of a national study. Acta Paediatr 1992; 81:589-592 [Medline]
3. Stevens FM, Egan-Mitchell B, Cryan E, McCarthy CF, McNicholl B Decreasing incidence of coeliac disease. Arch Dis Child 1987; 62:465-468 [Abstract/Free Full Text]
4. Logan RFA, Rifkind EA, Busuttil A, Gilmour HM, Ferguson A Prevalence and "incidence" of celiac disease in Edinburgh and the Lothian region of Scotland. Gastroentereology 1986; 90:334-342 [Medline]
5. Greco L, Tozzi AE, Mayer M, Grimaldi M, Silano G, Auricchio S Unchanging clinical picture of coeliac disease presentation in Campania, Italy. Eur J Pediatr 1989; 148:610-613 [CrossRef][Medline]
6. Cavell B. Increased prevalences of coeliac disease in Sweden: revelence of changes in infant feeding practices. In: Auricchio S, Visakorpi J, eds. Common Food Intolerances 1: Epidemiology of Coeliac Disease. Basel, Sweden: Karger; 1992;2:25-44
7. Mäki M, Kallonen K, Lähdeaho ML, Visakorpi JK Changing pattern of childhood coeliac disease in Finland. Acta Paediatr Scand 1988; 77:408-412 [Medline]
8. Ceccarelli M, Caiulo VA, Ughi C Changing pattern of coeliac disease in Western Toscana . Acta Paediatr Scand 1991; 80:547-548 [Medline]
9. Ascher H, Holm K, Kristiansson B, Mäki M Different features of coeliac disease in two neighboring countries. Arch Dis Child 1993; 69:375-380 [Abstract/Free Full Text]
10. Weile B, Cavell B, Nivenius K, Krasilnikoff PA Striking differences in the incidence of childhood celiac disease between Denmark and Sweden: a plausible explanation. J Pediatr Gastroenterol Nutr 1995; 21:64-68 [Medline]
11. Catassi C, Fabiani E, Rätsch IM, The coeliac iceberg in Italy: a multicentre antgliadin antibodies screening for coeliac disease in school-age subjects. Acta Paediatr Suppl 1996; 412:29-35 [Medline]
12. Stern M, Teuscher M, Wechmann T Serologic screening for coeliac disease: methodological standards and quality control. Acta Paediatr Suppl 1996; 412:49-51 [Medline]
13. Grodzinsky E, Franzen L, Hed J, Ström M High prevalence of celiac disease in healthy adults revealed by antigliadin antibodies. Ann Allergy 1992; 69:66-70 [Medline]
14. Mäki M, Holm K, Lipsanen V, Serologic markers and HLA genes among healthy first-degree relatives of patients with coeliac disease. Lancet 1991; ii:1350-1353
15. Bonamico M, Mariani P, Mazzilli MC, Frequency and clinical pattern of celiac disease among siblings of celiac children. J Pediatr Gastroenterol Nutr 1996; 23:159-163 [CrossRef][Medline]
16. Carlsson A, Axelsson I, Borulf S, Prevalence of IgA-antigliadin antibodies and IgA-antiendomysium antibodies related to celiac disease in children with Down syndrome. Pediatrics 1998; 101:272-275 [Abstract/Free Full Text]
17. Sigurs N, Johansson C, Elfstrand PO, Viander M, Lanner A Prevalence of coeliac disease in diabetic children and adolescents in Sweden. Acta Paediatr 1993; 82:748-751 [Medline]
18. Carlsson A, Axelsson I, Borulf S, Prevalence of IgA-antigliadin antibodies and IgA-antiendomysium antibodies at diagnosis of insulin-dependent diabetes mellitus (IDDM) in Swedish children and adolescents. Pediatrics 1999; 103:1248-1258 [Abstract/Free Full Text]
19. Catassi C, Rätsch IM, Fabiani E, Coeliac disease in the year 2000: exploring the iceberg. Lancet 1994; i:200-203 [CrossRef]
20. National Board of Health and Welfare. Official Report: Public Health in Sweden 1977. Stockholm, Sweden: National Board of Health and Welfare; 1979
21. Working Group of European Society of Paediatric Gastroenterology and Nutrition Revised criteria for diagnosis of coeliac disease: report of Working Group of European Society of Paediatric Gastroenterology and Nutrition . Arch Dis Child 1990; 65:909-911 [Free Full Text]
22. Lerner A, Kumar V, Iancu TC Immunological diagnosis of childhood coeliac disease: comparison between antigliadin, antireticulin and antiendomysial antibodies. Clin Exp Immunol 1994; 95:78-82 [Medline]
23. Ascher H, Hahn-Zoric M, Hansson LÅ, Kilander AF, Nilsson LÅ, Tlaskalová H Value of serologic markers for clinical diagnosis and population studies of coeliac disease. Scand J Gastroenterol 1996; 31:61-67 [Medline]
24. Johnston SD, Watson RGP, McMillian SA, Sloan J, Love AHG Prevalence of coeliac disease in Northern Ireland. Lancet 1997; ii:1370
25. Catassi C, Rätsch IM, Gandolfi L Why is coeliac disease endemic in the people of the Sahara? Lancet 1999; ii:647-648